1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
11 PP/02 Harmonised TLS library version reporting; only show if debugging.
12 Layout now matches that introduced for other libraries in 4.74 PP/03.
14 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
16 PP/04 New "dns_use_edns0" global option.
18 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
21 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
22 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
24 PP/07 Catch divide-by-zero in ${eval:...}.
31 NM/01 Workround for PCRE version dependency in version reporting
34 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
35 This fixes portability to compilers other than gcc, notably
36 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
38 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
39 makefiles for portability to HP-UX and POSIX correctness.
41 PP/01 Permit LOOKUP_foo enabling on the make command-line.
42 Also via indented variable definition in the Makefile.
43 (Debugging by Oliver Heesakkers).
45 PP/02 Restore caching of spamd results with expanded spamd_address.
46 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
48 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
49 Improves build reliability. Fix from: Frank Elsner
51 NM/02 Fix wide character breakage in the rfc2047 coding
52 Fixes bug 1064. Patch from Andrey N. Oktyabrski
54 NM/03 Allow underscore in dnslist lookups
55 Fixes bug 1026. Patch from Graeme Fowler
57 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
58 Code patches from Adam Ciarcinski of NetBSD.
60 NM/04 Fixed exiqgrep to cope with mailq missing size issue
63 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
64 is logged, to avoid truncation. Patch from John Horne.
66 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
67 Patch from Jakob Hirsch.
69 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
70 SQL string expansion failure details.
71 Patch from Andrey Oktyabrski.
73 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
74 Patch from Simon Arlott.
76 PP/09 New lookups functionality failed to compile on old gcc which rejects
77 extern declarations in function scope.
78 Patch from Oliver Fleischmann
80 PP/10 Use sig_atomic_t for flags set from signal handlers.
81 Check getgroups() return and improve debugging.
82 Fixed developed for diagnosis in bug 927 (which turned out to be
85 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
86 Patch from Mark Zealey.
88 PP/12 Bugzilla 1056: Improved spamd server selection.
89 Patch from Mark Zealey.
91 PP/13 Bugzilla 1086: Deal with maildir quota file races.
92 Based on patch from Heiko Schlittermann.
94 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
95 Patch from Uwe Doering, sign-off by Michael Haardt.
97 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
98 variable declaration deep within a block. Bug and patch from
101 PP/15 lookups-Makefile IRIX compatibilty coercion.
103 PP/16 Make DISABLE_DKIM build knob functional.
105 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
106 Patch by Simon Arlott
108 TF/03 Fix valgrind.h portability to C89 compilers that do not support
109 variable argument macros. Our copy now differs from upstream.
115 TF/01 Failure to get a lock on a hints database can have serious
116 consequences so log it to the panic log.
118 TF/02 Log LMTP confirmation messages in the same way as SMTP,
119 controlled using the smtp_confirmation log selector.
121 TF/03 Include the error message when we fail to unlink a spool file.
123 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
124 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
125 for maintaining out-of-tree patches for some time.
127 PP/01 Bugzilla 139: Documentation and portability issues.
128 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
129 Handle per-OS dynamic-module compilation flags.
131 PP/02 Let /dev/null have normal permissions.
132 The 4.73 fixes were a little too stringent and complained about the
133 permissions on /dev/null. Exempt it from some checks.
134 Reported by Andreas M. Kirchwitz.
136 PP/03 Report version information for many libraries, including
137 Exim version information for dynamically loaded libraries. Created
138 version.h, now support a version extension string for distributors
139 who patch heavily. Dynamic module ABI change.
141 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
142 privilege escalation vulnerability whereby the Exim run-time user
143 can cause root to append content of the attacker's choosing to
146 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
149 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
150 If dropping privileges for untrusted macros, we disabled normal logging
151 on the basis that it would fail; for the Exim run-time user, this is not
152 the case, and it resulted in successful deliveries going unlogged.
153 Fixed. Reported by Andreas Metzler.
159 PP/01 Date: & Message-Id: revert to normally being appended to a message,
160 only prepend for the Resent-* case. Fixes regression introduced in
161 Exim 4.70 by NM/22 for Bugzilla 607.
163 PP/02 Include check_rfc2047_length in configure.default because we're seeing
164 increasing numbers of administrators be bitten by this.
166 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
168 PP/03 Bugzilla 994: added openssl_options main configuration option.
170 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
172 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
174 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
176 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
177 without a peer certificate, leading to a segfault because of an
178 assumption that peers always have certificates. Be a little more
179 paranoid. Problem reported by Martin Tscholak.
181 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
182 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
183 NB: ClamAV planning to remove STREAM in "middle of 2010".
184 CL also introduces -bmalware, various -d+acl logging additions and
185 more caution in buffer sizes.
187 PP/09 Implemented reverse_ip expansion operator.
189 PP/10 Bugzilla 937: provide a "debug" ACL control.
191 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
193 PP/12 Bugzilla 973: Implement --version.
195 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
197 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
199 PP/15 Bugzilla 816: support multiple condition rules on Routers.
201 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
202 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
203 ignore trailing whitespace.
205 JJ/02 prevent non-panic DKIM error from being sent to paniclog
207 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
210 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
211 Notification from Dr Andrew Aitchison.
213 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
214 ExtendedDetectionInfo response format.
215 Notification from John Horne.
217 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
220 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
221 XSL and documented dependency on system catalogs, with examples of how
224 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
227 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
228 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
229 configuration file which is writeable by the Exim user or group.
231 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
232 of configuration files to cover files specified with the -C option if
233 they are going to be used with root privileges, not just the default
236 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
237 option (effectively making it always true).
239 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
240 files to be used while preserving root privileges.
242 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
243 that rogue child processes cannot use them.
245 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
246 run-time user, instead of root.
248 PP/28 Add WHITELIST_D_MACROS option to let some macros be overriden by the
249 Exim run-time user without dropping privileges.
251 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
252 result string, instead of calling string_vformat() twice with the same
255 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
256 for other users. Others should always drop root privileges if they use
257 -C on the command line, even for a whitelisted configure file.
259 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
261 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
267 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
268 $data_path, and $header_path variables; fixed documentation bugs and
271 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
272 exipick to access non-standard spools, including the "frozen" queue
275 NM/01 Bugzilla 965: Support mysql stored procedures.
276 Patch from Alain Williams
278 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
280 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
281 Patch from Andreas Metzler
283 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
284 Patch from Kirill Miazine
286 NM/05 Bugzilla 671: Added umask to procmail example.
288 JJ/03 installed exipick 20100323.0, fixing doc bug
290 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
291 directory. Notification and patch from Dan Rosenberg.
293 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
295 TK/02 Improve log output when DKIM signing operation fails.
297 MH/01 Treat the transport option dkim_domain as a colon separated
298 list, not as a single string, and sign the message with each element,
299 omitting multiple occurences of the same signer.
301 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
302 Bugzilla 985, 986. Patch by Simon Arlott
304 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
305 Patch by Simon Arlott
307 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
308 MBX locking. Notification from Dan Rosenberg.
314 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
316 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
318 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
320 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
322 NM/04 Bugzilla 915: Flag broken perl installation during build.
328 TK/01 Added patch by Johannes Berg that expands the main option
329 "spamd_address" if it starts with a dollar sign.
331 TK/02 Write list of recipients to X-Envelope-Sender header when building
332 the mbox-format spool file for content scanning (suggested by Jakob
335 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
336 (http://www.dcc-servers.net/) support via dccifd. Activated by
337 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
339 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
340 by Mark Daniel Reidel <mr@df.eu>.
342 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
343 When building exim an external PCRE library is now needed -
344 PCRE is a system library on the majority of modern systems.
345 See entry on PCRE_LIBS in EDITME file.
347 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
348 conversation. Added nologin parameter to request.
349 Patch contributed by Kirill Miazine.
351 TF/01 Do not log submission mode rewrites if they do not change the address.
353 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
355 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
356 log files in place. Contributed by Roberto Lima.
358 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
360 TF/03 Bugzilla 615: When checking the local_parts router precondition
361 after a local_part_suffix or local_part_prefix option, Exim now
362 does not use the address's named list lookup cache, since this
363 contains cached lookups for the whole local part.
365 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
366 Robert Millan. Documentation is in experimental-spec.txt.
368 TF/04 Bugzilla 668: Fix parallel build (make -j).
370 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
372 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
373 Patch provided by Jan Srzednicki.
375 TF/05 Leading white space used to be stripped from $spam_report which
376 wrecked the formatting. Now it is preserved.
378 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
379 that they are available at delivery time.
381 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
383 TF/08 TLS error reporting now respects the incoming_interface and
384 incoming_port log selectors.
386 TF/09 Produce a more useful error message if an SMTP transport's hosts
387 setting expands to an empty string.
389 NM/06 Bugzilla 744: EXPN did not work under TLS.
390 Patch provided by Phil Pennock.
392 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
393 Patch provided by Richard Godbee.
395 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
396 acl_smtp_notquit, added index entry.
398 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
399 Patch provided by Eugene Bujak.
401 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
402 accept(). Patch provided by Maxim Dounin.
404 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
405 Patch provided by Phil Pennock.
407 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
409 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
410 Patch provided by Brad "anomie" Jorsch.
412 NM/14 Bugzilla 622: Added timeout setting to transport filter.
413 Patch provided by Dean Brooks.
415 TK/05 Add native DKIM support (does not depend on external libraries).
417 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
418 Patch provided by Graeme Fowler.
420 NM/16 Bugzilla 851: Documentation example syntax fix.
422 NM/17 Changed NOTICE file to remove references to embedded PCRE.
424 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
427 NM/19 Bugzilla 745: TLS version reporting.
428 Patch provided by Phil Pennock.
430 NM/20 Bugzilla 167: bool: condition support.
431 Patch provided by Phil Pennock.
433 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
434 clients. Patch provided by Phil Pennock.
436 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
437 Patch provided by Brad "anomie" Jorsch.
439 NM/23 Bugzilla 687: Fix misparses in eximstats.
440 Patch provided by Heiko Schlittermann.
442 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
443 Patch provided by Heiko Schlittermann.
445 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
446 plus update to original patch.
448 NM/26 Bugzilla 799: Documentation correction for ratelimit.
450 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
451 Patch provided by David Brownlee.
453 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
455 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
457 NM/30 Bugzilla 888: TLS documentation bugfixes.
459 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
461 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
462 Unlike the original bugzilla I have changed all shell scripts in src tree.
464 NM/33 Bugzilla 898: Transport filter timeout fix.
465 Patch by Todd Rinaldo.
467 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
468 Patch by Serge Demonchaux.
470 NM/35 Bugzilla 39: Base64 decode bug fixes.
471 Patch by Jakob Hirsch.
473 NM/36 Bugzilla 909: Correct connect() call in dcc code.
475 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
477 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
479 NM/39 Bugzilla 911: Fixed MakeLinks build script.
485 TK/01 Add preliminary DKIM support. Currently requires a forked version of
486 ALT-N's libdkim that I have put here:
487 http://duncanthrax.net/exim-experimental/
489 Note to Michael Haardt: I had to rename some vars in sieve.c. They
490 were called 'true' and it seems that C99 defines that as a reserved
491 keyword to be used with 'bool' variable types. That means you could
492 not include C99-style headers which use bools without triggering
493 build errors in sieve.c.
495 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
496 as mailq or other aliases. Changed the --help handling significantly
497 to do whats expected. exim_usage() emits usage/help information.
499 SC/01 Added the -bylocaldomain option to eximstats.
501 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
503 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
505 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
511 PH/01 Another patch from the Sieve maintainer.
513 PH/02 When an IPv6 address is converted to a string for single-key lookup
514 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
515 dots are used instead of colons so that keys in lsearch files need not
516 contain colons. This was done some time before quoting was made available
517 in lsearch files. However, iplsearch files do require colons in IPv6 keys
518 (notated using the quote facility) so as to distinguish them from IPv4
519 keys. This meant that lookups for IP addresses in host lists did not work
520 for iplsearch lookups.
522 This has been fixed by arranging for IPv6 addresses to be expressed with
523 colons if the lookup type is iplsearch. This is not incompatible, because
524 previously such lookups could never work.
526 The situation is now rather anomolous, since one *can* have colons in
527 ordinary lsearch keys. However, making the change in all cases is
528 incompatible and would probably break a number of configurations.
530 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
533 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
534 conversion specification without a maximum field width, thereby enabling
535 a rogue spamd server to cause a buffer overflow. While nobody in their
536 right mind would setup Exim to query an untrusted spamd server, an
537 attacker that gains access to a server running spamd could potentially
538 exploit this vulnerability to run arbitrary code as the Exim user.
540 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
541 $primary_hostname instead of what libspf2 thinks the hosts name is.
543 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
544 a directory entry by the name of the lookup key. Previously, if a
545 symlink pointed to a non-existing file or a file in a directory that
546 Exim lacked permissions to read, a lookup for a key matching that
547 symlink would fail. Now it is enough that a matching directory entry
548 exists, symlink or not. (Bugzilla 503.)
550 PH/03 The body_linecount and body_zerocount variables are now exported in the
553 PH/04 Added the $dnslist_matched variable.
555 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
556 This means they are set thereafter only if the connection becomes
559 PH/06 Added the client_condition to authenticators so that some can be skipped
560 by clients under certain conditions.
562 PH/07 The error message for a badly-placed control=no_multiline_responses left
563 "_responses" off the end of the name.
565 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
567 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
568 (without spaces) instead of just copying the configuration text.
570 PH/10 Added the /noupdate option to the ratelimit ACL condition.
572 PH/11 Added $max_received_linelength.
574 PH/12 Added +ignore_defer and +include_defer to host lists.
576 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
577 way in which PCRE > 7.0 is built.
579 PH/14 Implemented queue_only_load_latch.
581 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
582 MAIL command. The effect was to mangle the value on 64-bit systems.
584 PH/16 Another patch from the Sieve maintainer.
586 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
588 PH/18 If a system quota error occurred while trying to create the file for
589 a maildir delivery, the message "Mailbox is full" was not appended to the
590 bounce if the delivery eventually timed out. Change 4.67/27 below applied
591 only to a quota excession during the actual writing of the file.
593 PH/19 It seems that peer DN values may contain newlines (and other non-printing
594 characters?) which causes problems in log lines. The DN values are now
595 passed through string_printing() before being added to log lines.
597 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
598 and InterBase are left for another time.)
600 PH/21 Added message_body_newlines option.
602 PH/22 Guard against possible overflow in moan_check_errorcopy().
604 PH/23 POSIX allows open() to be a macro; guard against that.
606 PH/24 If the recipient of an error message contained an @ in the local part
607 (suitably quoted, of course), incorrect values were put in $domain and
608 $local_part during the evaluation of errors_copy.
614 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
615 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
618 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
619 issue a MAIL command.
621 PH/02 In an ACL statement such as
623 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
625 if a client was not listed at all, or was listed with a value other than
626 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
627 the condition was not true (as it should be), so access was not denied.
628 The bug was that the ! inversion was incorrectly passed on to the second
629 item. This has been fixed.
631 PH/03 Added additional dnslists conditions == and =& which are different from
632 = and & when the dns lookup returns more than one IP address.
634 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
635 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
637 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
638 FSYNC, which compiles an option called disable_fsync that allows for
639 bypassing fsync(). The documentation is heavily laced with warnings.
641 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
643 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
644 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
645 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
646 including adding "make clean"; (3) Added -fPIC when compiling the test
647 dynamically loaded module, to get rid of a warning.
649 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
650 message fails, move_frozen_messages = true and ignore_bounce_errors_after
651 = 0s. The bug is otherwise harmless.
653 PH/07 There was a bug in the dovecot authenticator such that the value of
654 $auth1 could be overwritten, and so not correctly preserved, after a
655 successful authentication. This usually meant that the value preserved by
656 the server_setid option was incorrect.
658 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
660 PH/09 Installed PCRE release 7.0.
662 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
663 run for batched SMTP input. It is now run at the start of every message
664 in the batch. While fixing this I discovered that the process information
665 (output by running exiwhat) was not always getting set for -bs and -bS
666 input. This is fixed, and it now also says "batched" for BSMTP.
668 PH/11 Added control=no_pipelining.
670 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
671 patch, slightly modified), and move the expansion of helo_data till after
672 the connection is made in the smtp transport (so it can use these
675 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
677 PH/14 Added log_selector = +pid.
679 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
681 PH/16 Add ${if forany and ${if forall.
683 PH/17 Added dsn_from option to vary the From: line in DSNs.
685 PH/18 Flush SMTP output before performing a callout, unless control =
686 no_callout_flush is set.
688 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
689 was true (the default) a successful delivery failed to delete the retry
690 item, thus causing premature timeout of the address. The bug is now
693 PH/20 Added hosts_avoid_pipelining to the smtp transport.
695 PH/21 Long custom messages for fakedefer and fakereject are now split up
696 into multiline reponses in the same way that messages for "deny" and
697 other ACL rejections are.
699 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
700 with slight modification.
702 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
703 draft, changing the syntax and factoring some duplicate code".
705 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
706 for deliveries of the second and subsequent messages over the same SMTP
709 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
710 ${reduce, with only minor "tidies".
712 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
714 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
715 expansion side effects.
717 PH/27 When a message times out after an over-quota error from an Exim-imposed
718 quota, the bounce message says "mailbox is full". This message was not
719 being given when it was a system quota that was exceeded. It now should
722 MH/03 Made $recipients available in local_scan(). local_scan() already has
723 better access to the recipient list through recipients_list[], but
724 $recipients can be useful in postmaster-provided expansion strings.
726 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
727 in the case of a MAIL command with additional options following the
728 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
729 were accidentally chopped off.
731 PH/29 SMTP synchronization checks are implemented when a command is read -
732 there is a check that no more input is waiting when there shouldn't be
733 any. However, for some commands, a delay in an ACL can mean that it is
734 some time before the response is written. In this time, more input might
735 arrive, invalidly. So now there are extra checks after an ACL has run for
736 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
737 pipelining has not been advertised.
739 PH/30 MH's patch to allow iscntrl() characters to be list separators.
741 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
742 returned in the SMTP response when smtp_return_error_details was false.
745 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
746 instead of the C I/O that was originally supplied, because problems were
749 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
750 Exim which did not show up earlier: it was assuming that a call to
751 SSL_CTX_set_info_callback() might give an error value. In fact, there is
752 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
753 was a macro that became an assignment, so it seemed to work. This has
754 changed to a proper function call with a void return, hence the compile
755 error. Exim's code has been fixed.
757 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
760 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
762 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
764 PH/37 If a message is not accepted after it has had an id assigned (e.g.
765 because it turns out to be too big or there is a timeout) there is no
766 "Completed" line in the log. When some messages of this type were
767 selected by exigrep, they were listed as "not completed". Others were
768 picked up by some special patterns. I have improved the selection
769 criteria to be more general.
771 PH/38 The host_find_failed option in the manualroute router can now be set
772 to "ignore", to completely ignore a host whose IP address cannot be
773 found. If all hosts are ignored, the behaviour is controlled by the new
774 host_all_ignored option.
776 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
777 homing or because of multiple MX records with /mx) generated more than
778 one IP address, and the following item turned out to be the local host,
779 all the secondary addresses of the first item were incorrectly removed
780 from the list, along with the local host and any following hosts (which
781 is what is supposed to happen).
783 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
784 whoever called Exim into the -H file. In the case of the daemon it was
785 behaving confusingly. When first started, it used values for whoever
786 started the daemon, but after a SIGHUP it used the Exim user (because it
787 calls itself on a restart). I have changed the code so that it now always
790 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
791 message are rejected with the same error (e.g. no authentication or bad
792 sender address), and a DATA command is nevertheless sent (as can happen
793 with PIPELINING or a stupid MUA), the error message that was given to the
794 RCPT commands is included in the rejection of the DATA command. This is
795 intended to be helpful for MUAs that show only the final error to their
798 PH/42 Another patch from the Sieve maintainer.
800 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
801 Eximstats - Fixed some broken HTML links and added missing column headers
803 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
804 columns for Rejects, Temp Rejects, Ham, and Spam rows.
806 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
808 PH/43 Yet another patch from the Sieve maintainer.
810 PH/44 I found a way to check for a TCP/IP connection going away before sending
811 the response to the final '.' that terminates a message, but only in the
812 case where the client has not sent further data following the '.'
813 (unfortunately, this is allowed). However, in many cases there won't be
814 any further data because there won't be any more messages to send. A call
815 to select() can be used: if it shows that the input is "ready", there is
816 either input waiting, or the socket has been closed. An attempt to read
817 the next input character can distinguish the two cases. Previously, Exim
818 would have sent an OK response which the client would never have see.
819 This could lead to message repetition. This fix should cure that, at
820 least in a lot of common cases.
822 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
823 advertised in response to EHLO.
829 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
830 fixed by 4.65/MH/01 (is this a record?) are fixed:
832 (i) An empty string was always treated as zero by the numeric comparison
833 operators. This behaviour has been restored.
835 (ii) It is documented that the numeric comparison operators always treat
836 their arguments as decimal numbers. This was broken in that numbers
837 starting with 0 were being interpreted as octal.
839 While fixing these problems I realized that there was another issue that
840 hadn't been noticed. Values of message_size_limit (both the global option
841 and the transport option) were treated as octal if they started with 0.
842 The documentation was vague. These values are now always treated as
843 decimal, and I will make that clear in the documentation.
849 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
850 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
853 MH/01 Don't check that the operands of numeric comparison operators are
854 integers when their expansion is in "skipping" mode (fixes bug
855 introduced by 4.64-PH/07).
857 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
858 child addresses, Exim now panics and dies. Previously, because the count
859 is held in a short int, deliveries were likely to be lost. As such a
860 large number of recipients for a single message is ridiculous
861 (performance will be very, very poor), I have chosen to impose a limit
862 rather than extend the field.
868 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
869 leftover -K file (the existence of which was triggered by #402).
870 While we were at it, introduced process PID as part of the -K
871 filename. This should rule out race conditions when creating
874 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
875 processing considerably. Previous code took too long for large mails,
876 triggering a timeout which in turn triggers #401.
878 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
879 in the DK code in transports.c. sendfile() is not really portable,
880 hence the _LINUX specificness.
882 TF/01 In the add_headers option to the mail command in an Exim filter,
883 there was a bug that Exim would claim a syntax error in any
884 header after the first one which had an odd number of characters
887 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
888 callout verification, Exim cached a "reject" for the entire domain. This
889 is correct for most verifications, but it is not correct for a recipient
890 verification with use_sender or use_postmaster set, because in that case
891 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
892 case of MAIL FROM:<> rejection from other early rejections (e.g.
893 rejection of HELO). When verifying a recipient using a non-null MAIL
894 address, the cache is ignored if it shows MAIL FROM:<> rejection.
895 Whatever the result of the callout, the value of the domain cache is
896 left unchanged (for any other kind of callout, getting as far as trying
897 RCPT means that the domain itself is ok).
899 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
902 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
903 manifest itself as EPIPE rather than ECONNECT. When tidying away a
904 session, the daemon ignores ECONNECT errors and logs others; it now
905 ignores EPIPE as well.
907 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
908 (quoted-printable decoding).
910 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
911 later the small subsequent patch to fix an introduced bug.
913 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
915 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
917 PH/08 An error is now given if message_size_limit is specified negative.
919 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
920 to be given (somewhat) arbitrary names.
922 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
925 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
926 miscellaneous code fixes
928 PH/10 Added the log_reject_target ACL modifier to specify where to log
931 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
932 hostname. This is wrong, because it relates to the incoming message (and
933 probably the interface on which it is arriving) and not to the outgoing
934 callout (which could be using a different interface). This has been
935 changed to use the value of the helo_data option from the smtp transport
936 instead - this is what is used when a message is actually being sent. If
937 there is no remote transport (possible with a router that sets up host
938 addresses), $smtp_active_hostname is used.
940 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
941 tweaks were necessary in order to get it to work (see also 21 below):
942 (a) The code assumed that strncpy() returns a negative number on buffer
943 overflow, which isn't the case. Replaced with Exim's string_format()
945 (b) There were several signed/unsigned issues. I just did the minimum
946 hacking in of casts. There is scope for a larger refactoring.
947 (c) The code used strcasecmp() which is not a standard C function.
948 Replaced with Exim's strcmpic() function.
949 (d) The code set only $1; it now sets $auth1 as well.
950 (e) A simple test gave the error "authentication client didn't specify
951 service in request". It would seem that Dovecot has changed its
952 interface. Fortunately there's a specification; I followed it and
953 changed what the client sends and it appears to be working now.
955 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
958 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
959 address A is aliased to B and C, where B exists and C does not. Without
960 -v the output is "A verified" because verification stops after a
961 successful redirection if more than one address is generated. However,
962 with -v the child addresses are also verified. Exim was outputting "A
963 failed to verify" and then showing the successful verification for C,
964 with its parentage. It now outputs "B failed to verify", showing B's
965 parentage before showing the successful verification of C.
967 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
968 look up a TXT record in a specific list after matching in a combined
971 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
972 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
973 they consult the DNS. I had assumed they would set it the way they
974 wanted; and indeed my experiments on Linux seem to show that in some
975 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
976 To be on the safe side, however, I have now made the interface to
977 host_find_byname() similar to host_find_bydns(), with an argument
978 containing the DNS resolver options. The host_find_byname() function now
979 sets these options at its start, just as host_find_bydns() does. The smtp
980 transport options dns_qualify_single and dns_search_parents are passed to
981 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
982 of host_find_byname() use the default settings of RES_DEFNAMES
983 (qualify_single) but not RES_DNSRCH (search_parents).
985 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
986 spool_read_header() do less string testing, by means of a preliminary
987 switch on the second character of optional "-foo" lines. (This is
988 overdue, caused by the large number of possibilities that now exist.
989 Originally there were few.) While I was there, I also converted the
990 str(n)cmp tests so they don't re-test the leading "-" and the first
991 character, in the hope this might squeeze out yet more improvement.
993 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
994 flag allowing group syntax was set by the header_syntax check but not
995 turned off, possible causing trouble later; (2) The flag was not being
996 set at all for the header_verify test, causing "group"-style headers to
997 be rejected. I have now set it in this case, and also caused header_
998 verify to ignore an empty address taken from a group. While doing this, I
999 came across some other cases where the code for allowing group syntax
1000 while scanning a header line wasn't quite right (mostly, not resetting
1001 the flag correctly in the right place). These bugs could have caused
1002 trouble for malformed header lines. I hope it is now all correct.
1004 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1005 with the "reply" argument non-NULL. The code, however (which originally
1006 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1007 but it didn't always do it. This confused somebody who was copying the
1008 code for some other use. I have removed all the tests.
1010 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1011 feature that was used to support insecure browsers during the U.S. crypto
1012 embargo. It requires special client support, and Exim is probably the
1013 only MTA that supported it -- and would never use it because real RSA is
1014 always available. This code has been removed, because it had the bad
1015 effect of slowing Exim down by computing (never used) parameters for the
1016 RSA_EXPORT functionality.
1018 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1019 authenticator to fail if there's a tab character in the incoming data
1020 (there should never be unless someone is messing about, as it's supposed
1021 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1022 if the connection is using TLS or if the remote IP is the same as the
1023 local IP, and the "valid-client-cert option" if a client certificate has
1026 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1027 authenticators. This can be used for authorization after authentication
1028 succeeds. (In the case of plaintext, it servers for both authentication
1031 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1032 if any retry times were supplied.
1034 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1035 connection, where there is no client IP address to check. In this
1036 situation, the verify now always succeeds.
1038 PH/25 Applied John Jetmore's -Mset patch.
1040 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1042 PH/27 In a string expansion for a processed (not raw) header when multiple
1043 headers of the same name were present, leading whitespace was being
1044 removed from all of them, but trailing whitespace was being removed only
1045 from the last one. Now trailing whitespace is removed from each header
1046 before concatenation. Completely empty headers in a concatenation (as
1047 before) are ignored.
1049 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1050 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1052 PH/29 [Removed. This was a change that I later backed out, and forgot to
1053 correct the ChangeLog entry (that I had efficiently created) before
1054 committing the later change.]
1056 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1057 address errors (4xx response to RCPT) over the same connection as other
1058 messages routed to the same hosts. Such deliveries are always "forced",
1059 so retry times are not inspected. This resulted in far too many retries
1060 for the affected addresses. The effect occurred only when there were more
1061 hosts than the hosts_max_try setting in the smtp transport when it had
1062 the 4xx errors. Those hosts that it had tried were not added to the list
1063 of hosts for which the message was waiting, so if all were tried, there
1064 was no problem. Two fixes have been applied:
1066 (i) If there are any address or message errors in an SMTP delivery, none
1067 of the hosts (tried or untried) are now added to the list of hosts
1068 for which the message is waiting, so the message should not be a
1069 candidate for sending over the same connection that was used for a
1070 successful delivery of some other message. This seems entirely
1071 reasonable: after all the message is NOT "waiting for some host".
1072 This is so "obvious" that I'm not sure why it wasn't done
1073 previously. Hope I haven't missed anything, but it can't do any
1074 harm, as the worst effect is to miss an optimization.
1076 (ii) If, despite (i), such a delivery is accidentally attempted, the
1077 routing retry time is respected, so at least it doesn't keep
1078 hammering the server.
1080 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1081 in ${readsocket because some servers need this prod.
1083 PH/32 Added some extra debug output when updating a wait-xxx database.
1085 PH/33 The hint "could be header name not terminated by colon", which has been
1086 given for certain expansion errors for a long time, was not being given
1087 for the ${if def:h_colon_omitted{... case.
1089 PH/34 The spec says: "With one important exception, whenever a domain list is
1090 being scanned, $domain contains the subject domain." There was at least
1091 one case where this was not true.
1093 PH/35 The error "getsockname() failed: connection reset by peer" was being
1094 written to the panic log as well as the main log, but it isn't really
1095 panic-worthy as it just means the connection died rather early on. I have
1096 removed the panic log writing for the ECONNRESET error when getsockname()
1099 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1100 runs only) independently of the message's sender address. This meant
1101 that, if the 4xx error was in fact related to the sender, a different
1102 message to the same recipient with a different sender could confuse
1103 things. In particualar, this can happen when sending to a greylisting
1104 server, but other circumstances could also provoke similar problems.
1105 I have changed the default so that the retry time for these errors is now
1106 based a combination of the sender and recipient addresses. This change
1107 can be overridden by setting address_retry_include_sender=false in the
1110 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1111 remote server are returned as part of bounce messages. This was not
1112 happening for LMTP over a pipe (the lmtp transport), but now it is the
1113 same for both kinds of LMTP.
1115 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1116 in header lines that were in fact CNAMEs. This is no longer the case.
1118 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1119 and queue runs started by the daemon processed all messages. This has
1120 been fixed so that -R and -S can now usefully be given with -q<time>.
1122 PH/40 Import PCRE release 6.7 (fixes some bugs).
1124 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1126 PH/42 Give an error if -q is specified more than once.
1128 PH/43 Renamed the variables $interface_address and $interface_port as
1129 $received_ip_address and $received_port, to make it clear that these
1130 values apply to message reception, and not to the outgoing interface when
1131 a message is delivered. (The old names remain recognized, of course.)
1133 PH/44 There was no timeout on the connect() call when using a Unix domain
1134 socket in the ${readsocket expansion. There now is.
1136 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1137 be meaningful with "accept".
1139 SC/01 Eximstats V1.43
1140 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1142 SC/02 Eximstats V1.44
1143 Use a glob alias rather than an array ref in the generated
1144 parser. This improves both readability and performance.
1146 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1147 Collect SpamAssassin and rejection statistics.
1148 Don't display local sender or destination tables unless
1149 there is data to show.
1150 Added average volumes into the top table text output.
1152 SC/04 Eximstats V1.46
1153 Collect data on the number of addresses (recipients)
1154 as well as the number of messages.
1156 SC/05 Eximstats V1.47
1157 Added 'Message too big' to the list of mail rejection
1158 reasons (thanks to Marco Gaiarin).
1160 SC/06 Eximstats V1.48
1161 Mainlog lines which have GMT offsets and are too short to
1162 have a flag are now skipped.
1164 SC/07 Eximstats V1.49 (Alain Williams)
1165 Added the -emptyok flag.
1167 SC/08 Eximstats V1.50
1168 Fixes for obtaining the IP address from reject messages.
1170 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1171 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1172 whitesspace changes from 4.64-PH/27
1174 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1177 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1178 are found, allow negative numbers in numeric criteria)
1180 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1182 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1183 to match changes made in 4.64-PH/43
1185 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1187 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1188 "domain" is actually an IP address, and force a failure. This locks out
1189 those revolvers/nameservers that support "A-for-A" lookups, in
1190 contravention of the specifications.
1192 PH/48 When a host name was looked up from an IP address, and the subsequent
1193 forward lookup of the name timed out, the host name was left in
1194 $sender_host_name, contrary to the specification.
1196 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1197 restricted to single-key lookups, Exim was not diagnosing an error if
1198 * or *@ was used with a query-style lookup.
1200 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1202 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1203 long ago, but noone interested enough thought of it. Let's just say that
1204 the "1.1" means that there are some new functions that weren't there at
1205 some point in the past.
1207 PH/51 Error processing for expansion failure of helo_data from an smtp
1208 transport during callout processing was broken.
1210 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1211 tested/used via the -bh/-bhc/-bs options.
1213 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1214 bug, fixed in subsequent PCRE releases).
1216 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1217 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1219 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1225 SC/01 Use a glob alias rather than an array ref in eximstats generated
1226 parser. This improves both readability and performance.
1228 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1229 Don't display local sender or destination tables in eximstats unless
1230 there is data to show.
1231 Added average volumes into the eximstats top table text output.
1233 SC/03 Collect data on the number of addresses (recipients) as well
1234 as the number of messages in eximstats.
1236 TF/01 Correct an error in the documentation for the redirect router. Exim
1237 does (usually) call initgroups() when daemonizing.
1239 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1240 with consistent privilege compared to when running as a daemon.
1242 TF/03 Note in the spec that $authenticated_id is not set for local
1243 submissions from trusted users.
1245 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1246 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1248 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1249 by adding some example configuration directives to the default
1250 configuration file. A little bit of work is required to uncomment the
1251 directives and define how usernames and passwords are checked, but
1252 there is now a framework to start from.
1254 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1255 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1256 without this. I don't know how relevant this is to other LDAP libraries.
1258 PH/02 Add the verb name to the "unknown ACL verb" error.
1260 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1262 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1264 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1265 directory not expanded when it should be if an expanded home directory
1266 was set for the address (which is overridden by the transport).
1268 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1271 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1272 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1273 because it is too late at that time, and has no effect.
1275 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1276 security issue with \' (bugzilla #107). I could not use the
1277 PQescapeStringConn() function, because it needs a PGconn value as one of
1280 PH/08 When testing addresses using -bt, indicate those final addresses that
1281 are duplicates that would not cause an additional delivery. At least one
1282 person was confused, thinking that -bt output corresponded to deliveries.
1283 (Suppressing duplicates isn't a good idea as you lose the information
1284 about possibly different redirections that led to the duplicates.)
1286 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1287 systems where poll() doesn't work, in particular OS X.
1289 PH/10 Added more information to debugging output for retry time not reached.
1291 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1292 operations in malware.c.
1294 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1297 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1298 syslog_duplication set false, log lines that would normally be written
1299 both the the main log and to the reject log were not written to syslog at
1302 PH/14 In the default configuration, change the use of "message" in ACL warn
1303 statements to "add_header".
1305 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1306 not followed by a command (e.g. "seen endif").
1308 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1309 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1312 PH/17 Added extra conditions to the default value of delay_warning_condition
1316 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1317 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1318 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1321 The Auto-Submitted: and various List- headers are standardised, whereas I
1322 don't think Precedence: ever was.
1324 PH/18 Refactored debugging code in route_finduser() to show more information,
1325 in particular, the error code if getpwnam() issues one.
1327 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1328 This is apparently needed in addition to the PH/07 change above to avoid
1329 any possible encoding problems.
1331 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1332 but not after initializing Perl.
1334 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1335 output them only if debugging. By default they are written stderr,
1336 apparently, which is not desirable.
1338 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1341 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1344 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1346 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1347 authenticated or an ident call has been made. Suppress the default
1348 values for $authenticated_id and $authenticated_sender (but permit -oMai
1349 and -oMas) when testing with -bh.
1351 PH/24 Re-jigged the order of the tests in the default configuration so that the
1352 tests for valid domains and recipients precede the DNS black list and CSA
1353 tests, on the grounds that those ones are more expensive.
1355 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1356 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1357 This bug exists in every version of Exim that I still have, right back to
1360 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1361 However, an attempt to turn on case-sensitivity in a regex key by
1362 including (?-i) didn't work because the subject string was already
1363 lowercased, and the effects were non-intuitive. It turns out that a
1364 one-line patch can be used to allow (?-i) to work as expected.
1370 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1371 other effects) broke the use of negated acl sub-conditions.
1373 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1376 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1377 "Deny" causes Exim to reject the incoming connection with a 554 error.
1378 Unfortunately, if there is a major crisis, such as a disk failure,
1379 tcp-wrappers gives "deny", whereas what one would like would be some
1380 kind of temporary error. A kludge has been added to help with this.
1381 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1382 554 error is used if errno is still zero or contains ENOENT (which occurs
1383 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1386 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1388 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1389 errors. Otherwise a message that provokes a temporary error (when other
1390 messages do not) can cause a whole host to time out.
1392 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1393 addresses were routed directly to files or pipes from a redirect router.
1394 File deliveries just didn't batch; pipe deliveries might have suffered
1397 PH/06 A failure to get a lock for a hints database would erroneously always say
1398 "Failed to get write lock", even when it was really a read lock.
1400 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1401 of 0600. This has been changed to use the value of the lockfile_mode
1402 option (which defaults to 0600).
1404 PH/08 Applied small patch from the Sieve maintainer.
1406 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1407 folder from quota calculations, a direct delivery into this folder messed
1408 up the contents of the maildirsize file. This was because the regex was
1409 used only to exclude .Trash (or whatever) when the size of the mailbox
1410 was calculated. There was no check that a delivery was happening into an
1411 excluded directory. This bug has been fixed by ignoring all quota
1412 processing for deliveries into excluded directories.
1414 PH/10 Added the maildirfolder_create_regex option to appendfile.
1420 PH/01 The code for finding all the local interface addresses on a FreeBSD
1421 system running IPv6 was broken. This may well have applied to all BSD
1422 systems, as well as to others that have similar system calls. The broken
1423 code found IPv4 interfaces correctly, but gave incorrect values for the
1424 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1425 that it would not match correctly against @[] and not recognize the IPv6
1428 PH/02 The ipliteral router was not recognizing addresses of the form user@
1429 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1431 PH/03 Added disable_ipv6.
1433 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1434 decoded form, because it is most often used to construct To: headers
1435 lines in autoreplies, and the decoded form may well be syntactically
1436 invalid. However, $reply_address has leading white space removed, and all
1437 newlines turned into spaces so that the autoreply transport does not
1440 PH/05 If group was specified without a user on a router, and no group or user
1441 was specified on a transport, the group from the router was ignored.
1443 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1444 for visible compile-time settings that can be used to change these
1445 numbers, for those that want even more. Backwards compatibility with old
1446 spool files has been maintained. However, going back to a previous Exim
1447 release will lost any variables that are in spool files.
1449 PH/07 Two small changes when running in the test harness: increase delay when
1450 passing a TCP/IP connection to a new process, in case the original
1451 process has to generate a bounce, and remove special handling of
1452 127.0.0.2 (sic), which is no longer necessary.
1454 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1455 be the same on different OS.
1457 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1460 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1461 whether --show-vars was specified or not
1463 JJ/02 exipick: Added support for new ACL variable spool format introduced
1466 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1467 syntactically invalid From: or Reply-to: line, and a filter used this to
1468 generate an autoreply, and therefore failed to obtain an address for the
1469 autoreply, Exim could try to deliver to a non-existent relative file
1470 name, causing unrelated and misleading errors. What now happens is that
1471 it logs this as a hard delivery error, but does not attempt to create a
1474 PH/11 The exinext utility has a -C option for testing purposes, but although
1475 the given file was scanned by exinext itself; it wasn't being passed on
1476 when Exim was called.
1478 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1479 an end-of-file indication when reading a command response.
1481 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1482 compiled. In many other places in Exim, IPv6 addresses are always
1483 recognized, so I have changed this. It also means that IPv4 domain
1484 literals of the form [IPV4:n.n.n.n] are now always recognized.
1486 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1487 used if the router is not running as root, for example, when verifying at
1488 ACL time, or when using -bh. The debugging output from this situation was
1489 non-existent - all you got was a failure to exec. I have made two
1492 (a) Failures to set uid/gid, the current directory, or a process leader
1493 in a subprocess such as that created by queryprogram now generate
1494 suitable debugging ouput when -d is set.
1496 (b) The queryprogram router detects when it is not running as root,
1497 outputs suitable debugging information if -d is set, and then runs
1498 the subprocess without attempting to change uid/gid.
1500 PH/15 Minor change to Makefile for building test_host (undocumented testing
1503 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1504 additional section of a DNS packet that returns MX or SRV records.
1505 Instead, it always explicitly searches for A/AAAA records. This avoids
1506 major problems that occur when a DNS server includes only records of one
1507 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1508 fixed another bug: if SRV records were looked up and the corresponding
1509 address records were *not* found in the additional section, the port
1510 values from the SRV records were lost.
1512 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1513 using the correct key (the original address) when searching the retry
1514 rules in order to find which one to use for generating the retry hint.
1516 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1517 adding the default one. Similarly, if it contains a Reply-To: header, the
1518 errors_reply_to option, if set, is not used.
1520 PH/19 When calculating a retry time, Exim used to measure the "time since
1521 failure" by looking at the "first failed" field in the retry record. Now
1522 it does not use this if it is later than than the arrival time of the
1523 message. Instead it uses the arrival time. This makes for better
1524 behaviour in cases where some deliveries succeed, thus re-setting the
1525 "first failed" field. An example is a quota failure for a huge message
1526 when small messages continue to be delivered. Without this change, the
1527 "time since failure" will always be short, possible causing more frequent
1528 delivery attempts for the huge message than are intended.
1529 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1531 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1532 $1, $2, $3) because the numerical variables can be reset during some
1533 expansion items (e.g. "match"), thereby losing the authentication data.
1535 PH/21 Make -bV show the size of off_t variables so that the test suite can
1536 decide whether to run tests for quotas > 2G.
1538 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1539 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1540 filecount value is greater than 2G or if a quota value is greater than 2G
1541 on a system where the size of off_t is not greater than 4, a panic error
1544 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1545 never match. The debug and -bh output now contains an explicit error
1546 message indicating a malformed IPv4 address or mask.
1548 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1549 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1550 PH/23 above applies.
1552 PH/25 Do not write to syslog when running in the test harness. The only
1553 occasion when this arises is a failure to open the main or panic logs
1554 (for which there is an explicit test).
1556 PH/26 Added the /no_tell option to "control=freeze".
1558 PH/27 If a host name lookup failed very early in a connection, for example, if
1559 the IP address matched host_lookup and the reverse lookup yielded a name
1560 that did not have a forward lookup, an error message of the form "no IP
1561 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1562 could be logged. Now it outputs the IP address instead of "NULL".
1564 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1565 allows the sender and the authenticated sender to be set when
1566 submitting a message from within Exim. Since child_open_exim() is
1567 documented for local_scan(), the new function should be too.
1569 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1570 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1571 results in an empty string is now treated as unset.
1573 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1575 PH/31 Added sender_verify_fail logging option.
1577 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1578 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1579 tidied the source and removed it altogether.
1581 PH/33 When a queue run was abandoned because the load average was too high, a
1582 log line was always written; now it is written only if the queue_run log
1583 selector is set. In addition, the log line for abandonment now contains
1584 information about the queue run such as the pid. This is always present
1585 in "start" and "stop" lines but was omitted from the "abandon" line.
1587 PH/34 Omit spaces between a header name and the colon in the error message that
1588 is given when verify = headers_syntax fails (if there are lots of them,
1589 the message gets confusing).
1591 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1592 names, as there are now some PTR records that contain slashes. This check
1593 is only to protect against broken name servers that fall over on strange
1594 characters, so the fact that it applies to all lookups doesn't matter.
1596 PH/36 Now that the new test suite is complete, we can remove some of the
1597 special code in Exim that was needed for the old test suite. For example,
1598 sorting DNS records because real resolvers return them in an arbitrary
1599 order. The new test suite's fake resolver always returns records in the
1602 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1603 bounces) except when queue_only is set, to avoid logging races between
1604 the different processes.
1606 PH/38 Panic-die if .include specifies a non-absolute path.
1608 PH/39 A tweak to the "H" retry rule from its user.
1610 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1611 a label. They prevented compilation on older perls.
1613 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1614 a warning to be raised on newish perls.
1616 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1617 on queue. Changes to match documented behaviour of showing count of
1618 messages matching specified criteria.
1620 PH/40 Changed the default ident timeout from 30s to 5s.
1622 PH/41 Added support for the use of login_cap features, on those BSD systems
1623 that have them, for controlling the resources used by pipe deliveries.
1625 PH/42 The content-scanning code uses fopen() to create files in which to put
1626 message data. Previously it was not paying any attention to the mode of
1627 the files. Exim runs with umask(0) because the rest of the code creates
1628 files with open(), and sets the required mode explicitly. Thus, these
1629 files were ending up world-writeable. This was not a big issue, because,
1630 being within the spool directory, they were not world-accessible. I have
1631 created a function called modefopen, which takes an additional mode
1632 argument. It sets umask(777), creates the file, chmods it to the required
1633 mode, then resets the umask. All the relevant calls to fopen() in the
1634 content scanning code have been changed to use this function.
1636 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1637 to 24 hours. This avoids potential overflow problems when processing G
1638 and H retry rules. I suspect nobody ever tinkers with this value.
1640 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1642 PH/45 When the plaintext authenticator is running as a client, the server's
1643 challenges are checked to ensure they are valid base64 strings. By
1644 default, the authentication attempt is cancelled if an invalid string is
1645 received. Setting client_ignore_invalid_base64 true ignores these errors.
1646 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1647 they are received. Thus, the responses can be made to depend on the
1648 challenges. If an invalid string is ignored, an empty string is placed in
1651 PH/46 Messages that are created by the autoreply transport now contains a
1652 References: header, in accordance with RFCs 2822 and 3834.
1654 PH/47 Added authenticated_sender_force to the smtp transport.
1656 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1658 PH/49 Installed latest patch from the Sieve maintainer.
1660 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1661 was also set, the appendfile transport was unnecessarily scanning a
1662 directory of message files (e.g. for maildir delivery) to find the count
1663 of files (along with the size), even though it did not need this
1664 information. It now does the scan only if it needs to find either the
1665 size of the count of files.
1667 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1669 PH/52 Two bugs concerned with error handling when the smtp transport is
1672 (i) Exim was not creating retry information for temporary errors given
1673 for individual recipients after the DATA command when the smtp transport
1674 was used in LMTP mode. This meant that they could be retried too
1675 frequently, and not timed out correctly.
1677 (ii) Exim was setting the flag that allows error details to be returned
1678 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1679 recipients that were returned after the DATA command.
1681 PH/53 This is related to PH/52, but is more general: for any failing address,
1682 when detailed error information was permitted to be returned to the
1683 sender, but the error was temporary, then after the final timeout, only
1684 "retry timeout exceeded" was returned. Now it returns the full error as
1685 well as "retry timeout exceeded".
1687 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1688 do this, and (what is worse) MTAs that accept it.
1690 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1691 will now be deprecated.
1693 PH/56 New os.c-cygwin from the Cygwin maintainer.
1695 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1696 formats (previously only available in exim formats via -bpr, -bpru,
1697 and -bpra. Now also available in native and exiqgrep formats)
1699 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1700 with very large, slow to parse queues
1702 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1704 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1706 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1707 responses to authentication challenges, though it was showing the
1708 challenges; (ii) I've removed the CR characters from the debug output for
1711 PH/58 Allow for the insertion of a newline as well as a space when a string
1712 is turned into more than one encoded-word during RFC 2047 encoding. The
1713 Sieve code now uses this.
1715 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1716 data_4xx, lost_connection, tls_required.
1718 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1719 message was being sent as an SMTP response.
1721 PH/61 Add -l and -k options to exicyclog.
1723 PH/62 When verifying, if an address was redirected to one new address, so that
1724 verification continued, and the new address failed or deferred after
1725 having set something in $address_data, the value of $address_data was not
1726 passed back to the ACL. This was different to the case when no
1727 redirection occurred. The value is now passed back in both cases.
1729 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1730 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1731 use login_cap.h, so on its own it isn't the distinguishing feature. The
1732 new name refers directly to the setclassresources() function.
1734 PH/65 Added configuration files for NetBSD3.
1736 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1738 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1739 is preferred over IPv4.
1741 PH/68 The bounce_return_message and bounce_return_body options were not being
1742 honoured for bounces generated during the reception of non-SMTP messages.
1743 In particular, this applied to messages rejected by the ACL. This bug has
1744 been fixed. However, if bounce_return_message is true and bounce_return_
1745 body is false, the headers that are returned for a non-SMTP message
1746 include only those that have been read before the error was detected.
1747 (In the case of an ACL rejection, they have all been read.)
1749 PH/69 The HTML version of the specification is now built in a directory called
1750 spec_html instead of spec.html, because the latter looks like a path with
1751 a MIME-type, and this confuses some software.
1753 PH/70 Catch two compiler warnings in sieve.c.
1755 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1756 function verify_get_ident() calls ip_connect() to connect a socket, but
1757 if the "connect()" function timed out, ip_connect() used to close the
1758 socket. However, verify_get_ident() also closes the socket later, and in
1759 between Exim writes to the log, which may get opened at this point. When
1760 the socket was closed in ip_connect(), the log could get the same file
1761 descriptor number as the socket. This naturally causes chaos. The fix is
1762 not to close the socket in ip_connect(); the socket should be closed by
1763 the function that creates it. There was only one place in the code where
1764 this was missing, in the iplookup router, which I don't think anybody now
1765 uses, but I've fixed it anyway.
1767 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1768 well as to direct DNS lookups. Otherwise the handling of names in host
1769 lists is inconsistent and therefore confusing.
1775 PH/01 Two changes to the default runtime configuration:
1777 (1) Move the checks for relay_from_hosts and authenticated clients from
1778 after to before the (commented out) DNS black list checks.
1780 (2) Add control=submission to the relay_from_hosts and authenticated
1781 clients checks, on the grounds that messages accepted by these
1782 statements are most likely to be submissions.
1784 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1786 (1) Generate an error if the third argument for the ${prvs expansion is
1789 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1792 (3) Reset the variables that are obtained from the first argument of
1793 ${prvscheck and used in the second argument before leaving the code,
1794 because their memory is reclaimed, so using them afterwards may do
1797 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1798 one (it's much easier than Tom thought :-).
1800 (5) Because of (4), we can now allow for the use of $prvscheck_result
1801 inside the third argument.
1803 PH/03 For some reason, the default setting of PATH when running a command from
1804 a pipe transport was just "/usr/bin". I have changed it to
1807 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1808 anything to be listed in the output from -bV.
1810 PH/05 When a filter generated an autoreply, the entire To: header line was
1811 quoted in the delivery log line, like this:
1813 => >A.N.Other <ano@some.domain> <original@ddress> ...
1815 This has been changed so that it extracts the operative address. There
1816 may be more than one such address. If so, they are comma-separated, like
1819 => >ano@some.domain,ona@other.domain <original@ddress> ...
1821 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1822 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1823 not being looked up in the rDNS to get a host name, Exim was showing the
1824 IP address twice in Received: lines, even though the IP addresses were
1825 identical. For example:
1827 Received: from [1.2.3.4] (helo=[1.2.3.4])
1829 However, if the real host name was known, it was omitting the HELO data
1830 if it matched the actual IP address. This has been tidied up so that it
1831 doesn't show the same IP address twice.
1833 PH/07 When both +timestamp and +memory debugging was on, the value given by
1834 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1835 called by the debug printing, thereby overwriting the timestamp buffer.
1836 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1838 PH/08 When the original message was included in an autoreply transport, it
1839 always said "this is a copy of the message, including all the headers",
1840 even if body_only or headers_only was set. It now gives an appropriate
1843 PH/09 Applied a patch from the Sieve maintainer which:
1845 o fixes some comments
1846 o adds the (disabled) notify extension core
1847 o adds some debug output for the result of if/elsif tests
1848 o points to the current vacation draft in the documentation
1849 and documents the missing references header update
1853 o fixes a bug in processing the envelope test (when testing
1854 multiple envelope elements, the last element determinted the
1857 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1858 Electronic Mail") by including:
1860 Auto-submitted: auto-generated
1862 in the messages that it generates (bounce messages and others, such as
1863 warnings). In the case of bounce messages for non-SMTP mesages, there was
1864 also a typo: it was using "Auto_submitted" (underscore instead of
1865 hyphen). Since every message generated by Exim is necessarily in response
1866 to another message, thes have all been changed to:
1868 Auto-Submitted: auto-replied
1870 in accordance with these statements in the RFC:
1872 The auto-replied keyword:
1874 - SHOULD be used on messages sent in direct response to another
1875 message by an automatic process,
1877 - MUST NOT be used on manually-generated messages,
1879 - MAY be used on Delivery Status Notifications (DSNs) and Message
1880 Disposition Notifications (MDNs),
1882 - MUST NOT be used on messages generated by automatic or periodic
1883 processes, except for messages which are automatic responses to
1886 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1887 to the default Received: header definition.
1889 PH/12 Added log selector acl_warn_skipped (default on).
1891 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1892 variables because (a) they are in the wrong storage pool and (b) even if
1893 they were copied, it wouldn't work properly because of the caching.
1895 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1896 checking when decoding. Apparently there are clients that generate
1897 overlong encoded strings. Why am I not surprised?
1899 PH/15 If the first argument of "${if match_address" was not empty, but did not
1900 contain an "@" character, Exim crashed. Now it writes a panic log message
1901 and treats the condition as false.
1903 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1905 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1906 extension "envelope-auth". The code is finished and in agreement with
1907 other implementations, but there is no documentation so far and in fact,
1908 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1909 not changing the active code.
1911 Print executed "if" and "elsif" statements when debugging is used. This
1912 helps a great deal to understand what a filter does.
1914 Document more things not specified clearly in RFC3028. I had all this
1915 sorted out, when out of a sudden new issues came to my mind. Oops."
1917 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1920 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1921 canonical form (as documented). However, after a host name lookup from
1922 the IP address, check_host() was doing a simple string comparison with
1923 addresses acquired from the DNS when checking that the found name did
1924 have the original IP as one of its addresses. Since any found IPv6
1925 addresses are likely to be in abbreviated form, the comparison could
1926 fail. Luckily, there already exists a function for doing the comparison
1927 by converting both addresses to binary, so now that is used instead of
1928 the text comparison.
1930 PH/20 There was another similar case to PH/19, when a complete host name was
1931 given in a host list; looking up its IP address could give an abbreviated
1932 form, whereas the current host's name might or might not be abbreviated.
1933 The same fix has been applied.
1939 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1940 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1943 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1944 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1946 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1948 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1949 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1950 Also, for "Auto-Submitted", treat anything other than "no" as
1951 non-personal, in accordance with RFC 3834. (Previously it treated
1952 anything starting "auto-" as non-personal.)
1954 TF/01 The control=submission/name=... option had a problem with syntax
1955 errors if the name included a slash character. The /name= option
1956 now slurps the rest of the string, so it can include any characters
1957 but it must come last in the list of options (after /sender_retain
1960 PH/05 Some modifications to the interface to the fake nameserver for the new
1968 TK/01 Added the "success_on_redirect" address verification option. See
1969 NewStuff for rationale and an example.
1971 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1973 PH/02 Patch to exigrep to allow it to work on syslog lines.
1975 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1976 fread() to skip over the body file's header line, because in Cygwin the
1977 header line is locked and is inaccessible.
1979 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1980 co-exist for some time) to make it clear that it is the Exim ID that is
1981 referenced, not the Message-ID: header line.
1983 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1984 string_format() function, because snprintf() does not exist on all
1987 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1988 ${stat: expansion item.
1990 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1991 protocol synchronization error", to keep the pedants happy.
1993 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1994 well as for IRIX systems, when gcc is being used. See the host.c source
1997 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1999 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2002 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2003 transport and to the smtp transport in LMTP mode.
2005 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2007 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2008 run a filter in a subprocess. This could lead to confusion in subsequent
2009 lookups in the parent process. There should also be a search_tidyup() at
2010 the end of the subprocess.
2012 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2013 only if the host matched helo_try_verify_hosts, which caused the
2014 verification to occur when the EHLO/HELO command was issued. The ACL just
2015 tested the remembered result. Now, if a previous verification attempt has
2016 not happened, "verify = helo" does it there and then.
2018 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2020 TK/03 Fix log output including CR from clamd.
2022 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2023 exist provoked a memory error which could cause a segfault.
2025 PH/15 Installed PCRE 6.2
2027 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2029 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2030 of the problem. Specifically, suggested +O2 rather than +O1 for the
2033 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2035 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2036 record, the port was not correctly logged when the outgoing_port log
2037 selector was set (it logged the transort's default port).
2039 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2040 fallback_hosts, and "hosts" in the smtp transport.
2042 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2043 host errors such as "Connection refused".
2045 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2046 authentication with radiusclient 0.4.9:
2048 - Error returned from rc_read_config was caught wrongly
2049 - Username/password not passed on to radius server due to wrong length.
2051 The presumption is that some radiusclient API changes for 4.51/PH/17
2052 were not taken care of correctly. The code is still untested by me (my
2053 Linux distribution still has 0.3.2 of radiusclient), but it was
2054 contributed by a Radius user.
2056 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2057 expanding the "port" option of the smtp transport.
2059 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2060 while reading a MIME header. Thanks to Tom Hughes for a patch.
2062 PH/24 Include config.h inside local_scan.h so that configuration settings are
2065 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2066 that in an ACL for RCPT (for example), you can examine exactly what was
2069 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2070 commands, but it was not correctly comparing the address with the actual
2071 client host address. Thus, it would show the EHLO address in Received:
2072 header lines when this was not necessary.
2074 PH/27 Added the % operator to ${eval:}.
2076 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2077 it should be ignoring failures (because with -C, for example, it has lost
2078 privilege). It wasn't ignoring creation failures other than "already
2081 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2084 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2085 because an input line was too long, either on its own, or by virtue of
2086 too many continuations, the temporary file was not being removed, and the
2087 return code was incorrect.
2089 PH/31 Missing "BOOL" in function definition in filtertest.c.
2091 PH/32 Applied Sieve patches from the maintainer.
2093 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2095 PH/33 Added "verify = not_blind".
2097 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2098 Local/Makefile (with some defaults set). These are used in built scripts
2099 such as exicyclog, but they have never been used in the exim_install
2100 script (though there are many overriding facilities there). I have
2101 arranged that the exim_install script now takes note of these two
2104 PH/35 Installed configuration files for Dragonfly.
2106 PH/36 When a locally submitted message by a trusted user did not contain a
2107 From: header, and the sender address was obtained from -f or from an SMTP
2108 MAIL command, and the trusted user did not use -F to supply a sender
2109 name, $originator_name was incorrectly used when constructing a From:
2110 header. Furthermore, $originator_name was used for submission mode
2111 messages from external hosts without From: headers in a similar way,
2112 which is clearly wrong.
2114 PH/37 Added control=suppress_local_fixups.
2116 PH/38 When log_selector = +received_sender was set, and the addition of the
2117 sender made the log line's construction buffer exactly full, or one byte
2118 less than full, an overflow happened when the terminating "\n" was
2121 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2122 when the result of a list match is failure because a DNS lookup failed.
2124 PH/40 RM_COMMAND is now used in the building process.
2126 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2127 the "build-* directories that it finds.
2129 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2130 address in a domain literal was a prefix of an interface address.
2132 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2133 when verifying a sender address, unless rewrite_headers is false.
2135 PH/44 Wrote a long comment about why errors_to addresses are verified as
2136 recipients, not senders.
2138 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2139 the ratelimit ACL was added.
2141 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2143 PH/46 Added extra information about PostgreSQL errors to the error string.
2145 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2146 avoiding the need to install special zones in a real server. This is
2147 backwards compatible; if it can't find the fake resolver, it drops back.
2148 Thus, both old and new test suites can be run.
2150 TF/02 Added util/ratelimit.pl
2152 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2153 clock is set back in time.
2155 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2156 Candler <B.Candler@pobox.com>.
2158 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2159 OK for addresses that are the result of redirections.
2161 PH/48 A number of further additions for the benefit of the new test suite,
2162 including a fake gethostbyname() that interfaces to the fake DNS resolver
2165 TF/06 The fix for widen_domains has also been applied to qualify_single and
2166 search_parents which are the other dnslookup options that can cause
2169 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2172 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2174 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2175 service" if it encountered an MX record with an empty target hostname.
2176 The message is now "an MX or SRV record indicated no SMTP service".
2178 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2179 if the DNS of the sending site is misconfigured. This is quite a
2180 common situation. This change restores the behaviour of treating a
2181 helo verification defer as a failure.
2183 PH/51 If self=fail was set on a router, the bounce message did not include the
2184 actual error message.
2190 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2192 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2193 command itself ended in error, the underlying message about the transport
2194 filter timeout was being overwritten with the pipe command error. Now the
2195 underlying error message should be appended to the second error message.
2197 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2199 PH/02 Reduce the amount of output that "make" produces by default. Full output
2200 can still be requested.
2202 PH/03 The warning log line about a condition test deferring for a "warn" verb
2203 was being output only once per connection, rather than after each
2204 occurrence (because it was using the same function as for successful
2205 "warn" verbs). This seems wrong, so I have changed it.
2207 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2208 it should not have, which might have caused a crash in the right
2209 circumstances, but probably never did.
2211 PH/04 Installed a modified version of Tony Finch's patch to make submission
2212 mode fix the return path as well as the Sender: header line, and to
2213 add a /name= option so that you can make the user's friendly name appear
2216 TF/03 Added the control = fakedefer ACL modifier.
2218 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2219 Mark Lowes for thorough testing.
2221 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2223 TK/03 Merged latest SRS patch from Miles Wilton.
2225 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2226 with the definition in sysexits.h (which is #included earlier).
2227 Fortunately, Exim does not actually use EX_OK. The code used to try to
2228 preserve the sysexits.h value, by assumimg that macro definitions were
2229 scanned for macro replacements. I have been disabused of this notion,
2230 so now the code just undefines EX_OK before #including unistd.h.
2232 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2233 in the smtp transport. When a block could not be written in a single
2234 write() function, the timeout was being re-applied to each part-write.
2235 This seems wrong - if the receiver was accepting one byte at a time it
2236 would take for ever. The timeout is now adjusted when this happens. It
2237 doesn't have to be particularly precise.
2239 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2240 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2242 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2243 without a domain if the check to <postmaster@domain> fails.
2245 SC/01 Eximstats: added -xls and the ability to specify output files
2246 (patch written by Frank Heydlauf).
2248 SC/02 Eximstats: use FileHandles for outputing results.
2250 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2252 SC/04 Eximstats: fixed display of large numbers with -nvr option
2254 SC/05 Eximstats: fixed merging of reports with empty tables.
2256 SC/06 Eximstats: added the -include_original_destination flag
2258 SC/07 Eximstats: removed tabs and trailing whitespace.
2260 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2262 TK/06 MBOX spool code: Add real "From " MBOX separator line
2263 so the .eml file is really in mbox format (even though
2264 most programs do not really care). Patch from Alex Miller.
2266 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2267 The latter is generated from $received_to and is only set if the
2268 message has one envelope recipient. SA can use these headers,
2269 obviously out-of-the-box. Patch from Alex Miller.
2271 PH/08 The ${def test on a variable was returning false if the variable's
2272 value was "0", contrary to what the specification has always said!
2273 The result should be true unless the variable is empty.
2275 PH/09 The syntax error of a character other than { following "${if
2276 def:variable_name" (after optional whitespace) was not being diagnosed.
2277 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2278 accidental colon was present, for example, could give incorrect results.
2280 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2281 result is used (not including appendfile, where other changes are about
2284 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2285 This involved changing a lot of size variables from int to off_t. It
2286 should work with maildirs and everything.
2288 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2289 spamd dying while we are connected to it.
2291 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2294 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2295 function for BDB 4.3.
2297 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2299 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2300 the sources for PCRE have been re-arranged and more files are now
2303 PH/15 The code I had for printing potentially long long variables in PH/11
2304 above was not the best (it lost precision). The length of off_t variables
2305 is now inspected at build time, and an appropriate printing format (%ld
2306 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2307 to be "long long int" or "long int". This is needed for the internal
2308 formatting function string_vformat().
2310 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2311 the configuration file to be ":syslog", then the script "guesses" where
2312 the logs files are, rather than using the compiled in default. In our
2313 case the guess is not the same as the compiled default, so the script
2314 suddenly stopped working when I started to use syslog. The patch checks
2315 to see if log_file_path is "". If so, it attempts to read it from exim
2316 with no configuration file to get the compiled in version, before it
2317 falls back to the previous guessing code."
2319 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2320 implementing BATV in an Exim configuration. See NewStuff for the gory
2323 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2324 Makefile that are specific to HP-UX.
2326 PH/18 If the "use_postmaster" option was set for a recipient callout together
2327 with the "random" option, the postmaster address was used as the MAIL
2328 FROM address for the random test, but not for the subsequent recipient
2329 test. It is now used for both.
2331 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2332 patch removes a few documentation additions to RFC 3028, because the
2333 latest draft now contains them. It adds the new en;ascii-case comparator
2334 and a new error check for 8bit text in MIME parts. Comparator and
2335 require names are now matched exactly. I enabled the subaddress
2336 extension, but it is not well tested yet (read: it works for me)."
2338 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2339 rework some of the code of TK/09 above to avoid the hardwired use of
2340 "%lld" and "long long". Replaced the call to snprintf() with a call to
2343 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2344 records point to non-existent hosts", "retry timeout exceeded", and
2345 "retry time not reached for any host after a long failure period".
2347 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2348 experimental DomainKeys support:
2350 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2351 (2) On an error such as an illegally used "control", the wrong name for
2352 the control was given.
2354 These problems did NOT occur unless DomainKeys support was compiled.
2356 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2358 PH/24 Added ${if match_ip condition.
2360 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2361 databases so that it will be absolutely obvious if a crash occurs in the
2362 DB library. This is a regular occurrence (often caused by mis-matched
2365 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2366 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2367 picked up on a user's system that detects such things. There doesn't seem
2368 to be a gcc warning option for this - only an attribute that has to be
2369 put on the function's prototype. It seems that in Fedora Core 4 they have
2370 set this on a number of new functions. No doubt there will be more in due
2373 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2374 specify a transport. However, if an address that was verified by such a
2375 router was the subject of a callout, Exim crashed because it tried to
2376 read the rcpt_include_affixes from the non-existent transport. Now it
2377 just assumes that the setting of that option is false. This bug was
2378 introduced by 4.51/PH/31.
2380 PH/28 Changed -d+all to exclude +memory, because that information is very
2381 rarely of interest, but it makes the output a lot bigger. People tend to
2382 do -d+all out of habit.
2384 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2385 code in os-type was giving problems when libc.so lives in lib64, like on
2388 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2389 aren't the modern standard, and it seems that some systems' include files
2390 don't always have them. Exim was already checking for some of the newer
2391 ones like T_AAAA, and defining it itself. I've added checks for all the
2392 record types that Exim uses.
2394 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2395 not automatically generating a new one, as it is supposed to. This
2396 prevented TLS from working. If the file did exist, but contained invalid
2397 data, a new version was generated, as expected. It was only the case of a
2398 non-existent file that was broken.
2400 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2401 with a change in libdomainkeys > 0.64.
2403 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2404 from DNS. If the selector record carries the flag, it now has
2405 precedence over the domain-wide flag.
2407 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2409 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2410 the use of a port name that isn't defined in /etc/services) occurred, the
2411 message was deferred as in a normal delivery, and thus remained on the
2412 spool, instead of being failed because of the mua_wrapper setting. This
2413 is now fixed, and I tidied up some of the mua_wrapper messages at the
2416 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2417 the messages in a hash of arrays rather than using individual hashes.
2418 This is a bit cleaner and results in dramatic memory savings, albeit
2419 at a slight CPU cost.
2421 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2422 as requested by Marc Sherman.
2424 SC/10 Eximstats: added histograms for user specified patterns as requested
2427 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2429 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2430 fopen() in the content-scanning modules that did not already have it.
2436 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2437 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2439 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2441 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2443 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2444 to test Sieve filters that use "vacation".
2446 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2447 that changes the way the GnuTLS parameters are stored in the cache file.
2448 The new format can be generated externally. For backward compatibility,
2449 if the data in the cache doesn't make sense, Exim assumes it has read an
2450 old-format file, and it generates new data and writes a new file. This
2451 means that you can't go back to an older release without removing the
2454 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2455 work if there are any delivery delays because "one_time" forces the
2456 parent to be marked "delivered", so its unseen clone is never tried
2457 again. For this reason, Exim now forbids the simultaneous setting of
2460 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2461 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2462 entry if you want to know the details. The fix, however, neglected to
2463 consider the case when local delivery batching is involved. The test for
2464 "previously delivered" was not happening when checking to see if an
2465 address could be batched with a previous (undelivered) one; under
2466 certain circumstances this could lead to multiple deliveries to the same
2469 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2470 in its include files, and this causes problems building Exim.
2472 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2473 header_syntax) but Exim was just ignoring anything given after a slash.
2474 In particular, this caused confusion with an attempt to use "verify =
2475 reverse_host_lookup/defer_ok". An error is now given when options are
2476 supplied for verify items that do not have them. (Maybe reverse_host_
2477 lookup should have a defer_ok option, but that's a different point.)
2479 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2480 defined by RFC 821) to 2048, because there were problems with some AUTH
2481 commands, and RFC 1869 says the size should be increased for extended
2482 SMTP commands that take arguments.
2484 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2487 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2488 "unknown" error; now it says that the functionality isn't in the binary.
2490 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2491 an address' error message when a string expansion fails (syntax or
2492 whatever). Otherwise the password may appear in the log. Following change
2493 PH/42 below, there is no longer a chance of it appearing in a bounce
2496 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2498 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2499 of its addresses was ever tried. (Bugzilla bug #2.)
2501 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2502 the result incorrectly in the debug output. (It correctly added a newline
2503 to what was transported.)
2505 TF/01 Added $received_time.
2507 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2508 commented out examples of how to interface to a virus scanner and to
2509 SpamAssassin. Also added commented examples of av_scanner and
2510 spamd_address settings.
2512 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2513 and controls are allowed in which ACLs. There were a couple of minor
2514 errors. Some of the entries in the conditions table (which is a table of
2515 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2516 negation of where the condition IS allowed.
2518 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2520 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2521 header file does not have a version number, so I've had to invent a new
2522 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2523 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2524 radiusclient), but it was contributed by a Radius user.
2526 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2527 files or directories, for OpenSSL.
2529 PH/19 When an Exim process that is running as root has to create an Exim log
2530 file, it does so in a subprocess that runs as exim:exim so as to get the
2531 ownership right at creation (otherwise, other Exim processes might see
2532 the file with the wrong ownership). There was no test for failure of this
2533 fork() call, which would lead to the process getting stuck as it waited
2534 for a non-existent subprocess. Forks do occasionally fail when resources
2535 run out. I reviewed all the other calls to fork(); they all seem to check
2538 PH/20 When checking for unexpected SMTP input at connect time (before writing
2539 the banner), Exim was not dealing correctly with a non-positive return
2540 from the read() function. If the client had disconnected by this time,
2541 the result was a log entry for a synchronization error with an empty
2542 string after "input=" when read() returned zero. If read() returned -1
2543 (an event I could not check), uninitialized data bytes were printed.
2544 There were reports of junk text (parts of files, etc) appearing after
2547 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2549 PH/22 Added support for macro redefinition, and (re)definition in between
2550 driver and ACL definitions.
2552 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2553 forgetting to use the resulting value; it was using the unexpanded value.
2555 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2556 hadn't been configured. The fix is from Juergen Kreileder, who
2557 understands it better than I do:
2559 "Here's what I see happening with three configured cyrus_sasl
2560 authenticators configured (plain, login, cram-md5):
2562 On startup auth_cyrus_sasl_init() gets called for each of these.
2563 This means three calls to sasl_listmech() without a specified mech_list.
2564 => SASL tests which mechs of all available mechs actually work
2565 => three warnings about OTP not working
2566 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2568 With the patch, sasl_listmech() also gets called three times. But now
2569 SASL's mech_list option is set to the server_mech specified in the the
2570 authenticator. Or in other words, the answer from sasl_listmech()
2571 gets limited to just the mech you're testing for (which is different
2573 => the return list contains just 'plain' or 'login', 'cram-md5' or
2574 nothing depending on the value of ob->server_mech.
2576 I've just tested the patch: Authentication still works fine,
2577 unavailable mechs specified in the exim configuration are still
2578 caught, and the auth.log warnings about OTP are gone."
2580 PH/25 When debugging is enabled, the contents of the command line are added
2581 to the debugging output, even when log_selector=+arguments is not
2584 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2585 answer is "GNU", and only if the return is "GNU/something" is the answer
2588 PH/27 $acl_verify_message is now set immediately after the failure of a
2589 verification in an ACL, and so is available in subsequent modifiers. In
2590 particular, the message can be preserved by coding like this:
2592 warn !verify = sender
2593 set acl_m0 = $acl_verify_message
2595 Previously, $acl_verify_message was set only while expanding "message"
2596 and "log_message" when a very denied access.
2598 PH/28 Modified OS/os.c-Linux with
2600 -#ifndef OS_LOAD_AVERAGE
2601 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2603 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2604 nomenclature these days.)
2606 PH/29 Installed patch from the Sieve maintainer that adds the options
2607 sieve_useraddress and sieve_subaddress to the redirect router.
2609 PH/30 In these circumstances:
2610 . Two addresses routed to the same list of hosts;
2611 . First host does not offer TLS;
2612 . First host accepts first address;
2613 . First host gives temporary error to second address;
2614 . Second host offers TLS and a TLS session is established;
2615 . Second host accepts second address.
2616 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2617 and peerdn, if requested) that were in fact used only for the second
2620 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2621 attention to any local part prefix or suffix that was matched by the
2622 router that accepted the address. It now behaves in the same way as it
2623 does for delivery: the affixes are removed from the local part unless
2624 rcpt_include_affixes is set on the transport.
2626 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2627 timeout during the DATA phase of an incoming message.
2629 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2630 applied a patch sanctioned by the Sieve maintainer.
2632 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2633 the uid or gid is negative. A case of a negative gid caused this to be
2634 noticed. The fix allows for either to be negative.
2636 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2637 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2640 PH/36 The change PH/12 above was broken. Fixed it.
2642 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2643 the grounds that routing the same address twice would always produce the
2644 same answer. This might have been true once, but it is certainly no
2645 longer true now. Routing a child address may depend on the previous
2646 routing that produced that child. Some complicated redirection strategies
2647 went wrong when messages had multiple recipients, and made Exim's
2648 behaviour dependent on the order in which the addresses were given.
2650 I have moved the duplicate checking until after the routing is complete.
2651 Exim scans the addresses that are assigned to local and remote
2652 transports, and removes any duplicates. This means that more work will be
2653 done, as duplicates will always all be routed, but duplicates are
2654 presumably rare, so I don't expect this is of any significance.
2656 For deliveries to pipes, files, and autoreplies, the duplicate checking
2657 still happens during the routing process, since they are not going to be
2660 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2661 It corrects a timeout issue with spamd. This is Ian's comment: "The
2662 background is that sometimes spamd either never reads data from a
2663 connection it has accepted, or it never writes response data. The exiscan
2664 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2665 blindly assumes that writes won't block so it may never time out."
2667 PH/39 Allow G after quota size as well as K and M.
2669 PH/40 The value set for $authenticated_id in an authenticator may not contain
2670 binary zeroes or newlines because the value is written to log lines and
2671 to spool files. There was no check on this. Now the value is run through
2672 the string_printing() function so that such characters are converted to
2673 printable escape sequences.
2675 PH/41 $message_linecount is a new variable that contains the total number of
2676 lines in the message. Compare $body_linecount, which is the count for the
2679 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2680 bounce and delay warning messages, except in certain special cases, which
2683 (a) An SMTP error message from a remote host;
2684 (b) A message specified in a :fail: redirection;
2685 (c) A message specified in a "fail" command in a system filter;
2686 (d) A message specified in a FAIL return from the queryprogram router;
2687 (e) A message specified by the cannot_route_message router option.
2689 In these cases only, Exim does include the error details in bounce and
2690 warning messages. There are also a few cases where bland messages such
2691 as "unrouteable address" or "local delivery error" are given.
2693 PH/43 $value is now also set for the "else" part of a ${run expansion.
2695 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2696 being worked on, but at least Exim now implements the latest version to
2699 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2700 process to complete was treated as a delivery failure, a timeout while
2701 writing the message to the pipe was logged, but erroneously treated as a
2702 successful delivery. Such timeouts include transport filter timeouts. For
2703 consistency with the overall process timeout, these timeouts are now
2704 treated as errors, giving rise to delivery failures by default. However,
2705 there is now a new Boolean option for the pipe transport called
2706 timeout_defer, which, if set TRUE, converts the failures into defers for
2707 both kinds of timeout. A transport filter timeout is now identified in
2710 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2711 systems where "make" and "gmake" are different, calling "gmake" at top
2712 level broke things. I've arranged for the value of $(MAKE) to be passed
2713 from the Makefile to this script so that it can call the same version of
2717 A note about Exim versions 4.44 and 4.50
2718 ----------------------------------------
2720 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2721 changes of various kinds. As a consequence, a big documentation update was
2722 needed. This delayed the release for rather longer than seemed good, especially
2723 in the light of a couple of (minor) security issues. Therefore, the changes
2724 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2725 release. So 4.44 and 4.50 are in effect two different branches that both start
2728 I have left the 4.50 change log unchanged; it contains all the changes since
2729 4.43. The change log for 4.44 is below; many of its items are identical to
2730 those for 4.50. This seems to be the most sensible way to preserve the
2731 historical information.
2737 1. Minor wording change to the doc/README.SIEVE file.
2739 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2740 computation of the current number of files was incorrect.
2742 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2743 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2746 4. Give more explanation in the error message when the command for a transport
2747 filter fails to execute.
2749 5. There are several places where Exim runs a non-Exim command in a
2750 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2751 was being done only for the command run by the queryprogram router. It is
2752 now done for all such subprocesses. The other cases are: ${run, transport
2753 filters, and the commands run by the lmtp and pipe transports.
2755 6. Added CONFIGURE_GROUP build-time option.
2757 7. Some older OS have a limit of 256 on the maximum number of file
2758 descriptors. Exim was using setrlimit() to set 1000 as a large value
2759 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2760 systems. I've change it so that if it can't get 1000, it tries for 256.
2762 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2763 was an oversight, and furthermore, ever since the addition of extra
2764 controls (e.g. 4.43/32), the checks on when to allow different forms of
2765 "control" were broken. There should now be diagnostics for all cases when a
2766 control that does not make sense is encountered.
2768 9. Added the /retain_sender option to "control=submission".
2770 10. $recipients is now available in the predata ACL (oversight).
2772 11. Tidy the search cache before the fork to do a delivery from a message
2773 received from the command line. Otherwise the child will trigger a lookup
2774 failure and thereby defer the delivery if it tries to use (for example) a
2775 cached ldap connection that the parent has called unbind on.
2777 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2778 of $address_data from the recipient verification was clobbered by the
2779 sender verification.
2781 13. The value of address_data from a sender verification is now available in
2782 $sender_address_data in subsequent conditions in the ACL statement.
2784 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2786 15. Added a new option "connect=<time>" to callout options, to set a different
2789 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2790 was its contents. (It was OK if the option was not defined at all.)
2792 17. A "Completed" log line is now written for messages that are removed from
2793 the spool by the -Mrm option.
2795 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2796 information about exactly what failed.
2798 19. Added -dd to debug only the daemon process.
2800 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2801 handles timeouts, both on the server side and network timeouts. Renamed the
2802 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2804 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2805 It is now set to "smtps".
2807 22. $host_address is now set to the target address during the checking of
2808 ignore_target_hosts.
2810 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2811 being passed; this would have caused $sender_host_name to have been used if
2812 matching the list had actually called for a host name (not very likely,
2813 since this list is usually IP addresses). A host name is now passed as
2816 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2817 code that specifies a non-restarting handler (typically sigaction() in
2818 modern systems) in an attempt to fix a rare and obscure crash bug.
2820 25. Narrowed the window for a race in the daemon that could cause it to ignore
2821 SIGCHLD signals. This is not a major problem, because they are used only to
2822 wake it up if nothing else does.
2824 26. A malformed maildirsize file could cause Exim to calculate negative values
2825 for the mailbox size or file count. Odd effects could occur as a result.
2826 The maildirsize information is now recalculated if the size or filecount
2829 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2830 support for a long time. Removed HAVE_SYS_VFS_H.
2832 28. Installed the latest version of exipick from John Jetmore.
2834 29. In an address list, if the pattern was not a regular expression, an empty
2835 subject address (from a bounce message) matched only if the pattern was an
2836 empty string. Non-empty patterns were not even tested. This was the wrong
2837 because it is perfectly reasonable to use an empty address as part of a
2838 database query. An empty address is now tested by patterns that are
2839 lookups. However, all the other forms of pattern expect the subject to
2840 contain a local part and a domain, and therefore, for them, an empty
2841 address still always fails if the pattern is not itself empty.
2843 30. Exim went into a mad DNS loop when attempting to do a callout where the
2844 host was specified on an smtp transport, and looking it up yielded more
2845 than one IP address.
2847 31. Re-factored the code for checking spool and log partition space into a
2848 function that finds that data and another that does the check. The former
2849 is then used to implement four new variables: $spool_space, $log_space,
2850 $spool_inodes, and $log_inodes.
2852 32. The RFC2047 encoding function was originally intended for short strings
2853 such as real names; it was not keeping to the 75-character limit for
2854 encoded words that the RFC imposes. It now respects the limit, and
2855 generates multiple encoded words if necessary. To be on the safe side, I
2856 have increased the buffer size for the ${rfc2047: expansion operator from
2859 33. It is now permitted to omit both strings after an "if" condition; if the
2860 condition is true, the result is "true". As before, when the second string
2861 is omitted, a false condition yields an empty string. This makes it less
2862 cumbersome to write custom ACL and router conditions.
2864 34. Failure to deliver a bounce message always caused it to be frozen, even if
2865 there was an errors_to setting on the router. The errors_to setting is now
2868 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2869 canonical form (fully expanded) before being placed in
2870 $sender_host_address.
2872 36. The table in the code that translates DNS record types into text (T_A to
2873 "A" for instance) was missing entries for NS and CNAME. It is just possible
2874 that this could have caused confusion if both these types were looked up
2875 for the same domain, because the text type is used as part of Exim's
2876 per-process caching. But the chance of anyone hitting this buglet seems
2879 37. The dnsdb lookup has been extended in a number of ways.
2881 (1) There is a new type, "zns", which walks up the domain tree until it
2882 finds some nameserver records. It should be used with care.
2884 (2) There is a new type, "mxh", which is like "mx" except that it returns
2885 just the host names, not the priorities.
2887 (3) It is now possible to give a list of domains (or IP addresses) to be
2888 looked up. The behaviour when one of the lookups defers can be
2889 controlled by a keyword.
2891 (4) It is now possible to specify the separator character for use when
2892 multiple records are returned.
2894 38. The dnslists ACL condition has been extended: it is now possible to supply
2895 a list of IP addresses and/or domains to be looked up in a particular DNS
2898 39. Added log_selector=+queue_time_overall.
2900 40. When running the queue in the test harness, wait just a tad after forking a
2901 delivery process, to get repeatability of debugging output.
2903 41. Include certificate and key file names in error message when GnuTLS fails
2904 to set them up, because the GnuTLS error message doesn't include the name
2905 of the failing file when there is a problem reading it.
2907 42. Allow both -bf and -bF in the same test run.
2909 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2911 44. The "Exiscan patch" is now merged into the mainline Exim source.
2913 45. Sometimes the final signoff response after QUIT could fail to get
2914 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2915 < 0 before doing a fflush(). This bug looks as though it goes back to the
2916 introduction of TLS in release 3.20, but "sometimes" must have been rare
2917 because the tests only now provoked it.
2919 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2920 (this can affect the format of dates).
2922 47. exim_tidydb, when checking for the continued existence of a message for
2923 which it has found a message-specific retry record, was not finding
2924 messages that were in split spool directories. Consequently, it was
2925 deleting retry records that should have stayed in existence.
2927 48. Steve fixed some bugs in eximstats.
2929 49. The SPA authentication driver was not abandoning authentication and moving
2930 on to the next authenticator when an expansion was forced to fail,
2931 contradicting the general specification for all authenticators. Instead it
2932 was generating a temporary error. It now behaves as specified.
2934 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2935 (the order specifies the preference for clients). The order is now AES256,
2936 AES128, 3DES, ARCFOUR128.
2938 51. Small patch to Sieve code - explicitly set From: when generating an
2941 52. Exim crashed if a remote delivery caused a very long error message to be
2942 recorded - for instance if somebody sent an entire SpamAssassin report back
2943 as a large number of 550 error lines. This bug was coincidentally fixed by
2944 increasing the size of one of Exim's internal buffers (big_buffer) that
2945 happened as part of the Exiscan merge. However, to be on the safe side, I
2946 have made the code more robust (and fixed the comments that describe what
2949 53. Now that there can be additional text after "Completed" in log lines (if
2950 the queue_time_overall log selector is set), a one-byte patch to exigrep
2951 was needed to allow it to recognize "Completed" as not the last thing in
2954 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2955 patch that reportedly fixes this has been added. I am not expert enough to
2956 create a test for it. This is what the patch creator wrote:
2958 "I found a little strange behaviour of ldap code when working with
2959 Windows 2003 AD Domain, where users was placed in more than one
2960 Organization Units. When I tried to give exim partial DN, the exit code
2961 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2962 But simultaneously result of request was absolutely normal ldap result,
2963 so I produce this patch..."
2965 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2966 REFERENCE, so I have modified the code to exclude the patch when that macro
2969 55. Some experimental protocols are using DNS PTR records for new purposes. The
2970 keys for these records are domain names, not reversed IP addresses. The
2971 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2972 leaves it alone. Component reversal etc. now happens only for IP addresses.
2975 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2977 57. Double the size of the debug message buffer (to 2048) so that more of very
2978 long debug lines gets shown.
2980 58. The exicyclog utility now does better if the number of log files to keep
2981 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2983 59. Two changes related to the smtp_active_hostname option:
2985 (1) $smtp_active_hostname is now available as a variable.
2986 (2) The default for smtp_banner uses $smtp_active_hostname instead
2987 of $primary_hostname.
2989 60. The host_aton() function is supposed to be passed a string that is known
2990 to be a valid IP address. However, in the case of IPv6 addresses, it was
2991 not checking this. This is a hostage to fortune. Exim now panics and dies
2992 if the condition is not met. A case was found where this could be provoked
2993 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2994 components; fortuitously, this particular loophole had already been fixed
2995 by change 4.50/55 above.
2997 If there are any other similar loopholes, the new check in host_aton()
2998 itself should stop them being exploited. The report I received stated that
2999 data on the command line could provoke the exploit when Exim was running as
3000 exim, but did not say which command line option was involved. All I could
3001 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3002 running as the user.
3005 61. There was a buffer overflow vulnerability in the SPA authentication code
3006 (which came originally from the Samba project). I have added a test to the
3007 spa_base64_to_bits() function which I hope fixes it.
3010 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3011 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3013 63. The daemon start-up calls getloadavg() while still root for those OS that
3014 need the first call to be done as root, but it missed one case: when
3015 deliver_queue_load_max is set with deliver_drop_privilege. This is
3016 necessary for the benefit of the queue runner, because there is no re-exec
3017 when deliver_drop_privilege is set.
3019 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3020 This has been fixed.
3022 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3023 was in use, was not putting the data itself into the right store pool;
3024 consequently, it could be overwritten for a subsequent message in the same
3025 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3028 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3030 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3031 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3032 the function were treating the return as a boolean value, which happened to
3033 work because 0=false and not-0=true, but is not correct code.
3035 68. The host_aton() function was not handling scoped IPv6 addresses (those
3036 with, for example, "%eth0" on the end) correctly.
3038 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3039 negated items (that is, ~something) in unsigned ints. Some compilers
3040 apparently mutter when there is no cast.
3042 70. If an address verification called from an ACL failed, and did not produce a
3043 user-specific message (i.e. there was only a "system" message), nothing was
3044 put in $acl_verify_message. In this situation, it now puts the system
3047 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3048 session; change 4.31/43 added the unwanted input to the log line - except
3049 that it did not do this in the start of session case. It now does.
3051 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3052 This is wrong and can cause the other end to generate a synchronization
3053 error if it is another Exim or anything else that does the synchronization
3054 check. A QUIT command is no longer sent after a timeout.
3056 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3057 during host lookups.
3059 74. The defer_ok option of callout verification was not working if it was used
3060 when verifying addresses in header lines, that is, for this case:
3062 verify = header_sender/callout=defer_ok
3064 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3065 those file descriptors could be used for SMTP connections. If anything
3066 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3067 could be sent to the SMTP client, causing chaos. The daemon now opens
3068 stdin, stdout, and stderr to /dev/null when it puts itself into the
3071 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3072 log by default. The user can override this with suitable Perl magic.
3074 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3075 the log message when discard triggers, was not working for the DATA ACL or
3076 for the non-SMTP ACL.
3078 78. Error message wording change in sieve.c.
3080 79. If smtp_accept_max_per_host was set, the number of connections could be
3081 restricted to fewer than expected, because the daemon was trying to set up
3082 a new connection before checking whether the processes handling previous
3083 connections had finished. The check for completed processes is now done
3084 earlier. On busy systems, this bug wouldn't be noticed because something
3085 else would have woken the daemon, and it would have reaped the completed
3088 80. If a message was submitted locally by a user whose login name contained one
3089 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3090 It caused a spool format error. I have fixed the spool reading code. A
3091 related problem was that the "from" clause in the Received: line became
3092 illegal because of the space(s). It is now covered by ${quote_local_part.
3094 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3097 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3099 83. Patch from Sieve maintainer to fix unterminated string problem in
3100 "vacation" handling.
3102 84. Some minor changes to the Linux configuration files to help with other
3103 OS variants using glibc.
3105 85. One more patch for Sieve to update vacation handling to latest spec.
3108 ----------------------------------------------------
3109 See the note above about the 4.44 and 4.50 releases.
3110 ----------------------------------------------------
3116 1. Change 4.43/35 introduced a bug that caused file counts to be
3117 incorrectly computed when quota_filecount was set in an appendfile
3120 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3121 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3124 3. Give more explanation in the error message when the command for a transport
3125 filter fails to execute.
3127 4. There are several places where Exim runs a non-Exim command in a
3128 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3129 was being done only for the command run by the queryprogram router. It is
3130 now done for all such subprocesses. The other cases are: ${run, transport
3131 filters, and the commands run by the lmtp and pipe transports.
3133 5. Some older OS have a limit of 256 on the maximum number of file
3134 descriptors. Exim was using setrlimit() to set 1000 as a large value
3135 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3136 systems. I've change it so that if it can't get 1000, it tries for 256.
3138 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3139 was an oversight, and furthermore, ever since the addition of extra
3140 controls (e.g. 4.43/32), the checks on when to allow different forms of
3141 "control" were broken. There should now be diagnostics for all cases when a
3142 control that does not make sense is encountered.
3144 7. $recipients is now available in the predata ACL (oversight).
3146 8. Tidy the search cache before the fork to do a delivery from a message
3147 received from the command line. Otherwise the child will trigger a lookup
3148 failure and thereby defer the delivery if it tries to use (for example) a
3149 cached ldap connection that the parent has called unbind on.
3151 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3152 of $address_data from the recipient verification was clobbered by the
3153 sender verification.
3155 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3156 was its contents. (It was OK if the option was not defined at all.)
3158 11. A "Completed" log line is now written for messages that are removed from
3159 the spool by the -Mrm option.
3161 12. $host_address is now set to the target address during the checking of
3162 ignore_target_hosts.
3164 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3165 being passed; this would have caused $sender_host_name to have been used if
3166 matching the list had actually called for a host name (not very likely,
3167 since this list is usually IP addresses). A host name is now passed as
3170 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3171 code that specifies a non-restarting handler (typically sigaction() in
3172 modern systems) in an attempt to fix a rare and obscure crash bug.
3174 15. Narrowed the window for a race in the daemon that could cause it to ignore
3175 SIGCHLD signals. This is not a major problem, because they are used only to
3176 wake it up if nothing else does.
3178 16. A malformed maildirsize file could cause Exim to calculate negative values
3179 for the mailbox size or file count. Odd effects could occur as a result.
3180 The maildirsize information is now recalculated if the size or filecount
3183 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3184 support for a long time. Removed HAVE_SYS_VFS_H.
3186 18. Updated exipick to current release from John Jetmore.
3188 19. Allow an empty sender to be matched against a lookup in an address list.
3189 Previously the only cases considered were a regular expression, or an
3192 20. Exim went into a mad DNS lookup loop when doing a callout where the
3193 host was specified on the transport, if the DNS lookup yielded more than
3196 21. The RFC2047 encoding function was originally intended for short strings
3197 such as real names; it was not keeping to the 75-character limit for
3198 encoded words that the RFC imposes. It now respects the limit, and
3199 generates multiple encoded words if necessary. To be on the safe side, I
3200 have increased the buffer size for the ${rfc2047: expansion operator from
3203 22. Failure to deliver a bounce message always caused it to be frozen, even if
3204 there was an errors_to setting on the router. The errors_to setting is now
3207 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3208 canonical form (fully expanded) before being placed in
3209 $sender_host_address.
3211 24. Updated eximstats to version 1.33
3213 25. Include certificate and key file names in error message when GnuTLS fails
3214 to set them up, because the GnuTLS error message doesn't include the name
3215 of the failing file when there is a problem reading it.
3217 26. Expand error message when OpenSSL has problems setting up cert/key files.
3220 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3221 (this can affect the format of dates).
3223 28. exim_tidydb, when checking for the continued existence of a message for
3224 which it has found a message-specific retry record, was not finding
3225 messages that were in split spool directories. Consequently, it was
3226 deleting retry records that should have stayed in existence.
3228 29. eximstats updated to version 1.35
3229 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3230 1.35 - bugfix such that pie charts by volume are generated correctly
3232 30. The SPA authentication driver was not abandoning authentication and moving
3233 on to the next authenticator when an expansion was forced to fail,
3234 contradicting the general specification for all authenticators. Instead it
3235 was generating a temporary error. It now behaves as specified.
3237 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3238 (the order specifies the preference for clients). The order is now AES256,
3239 AES128, 3DES, ARCFOUR128.
3241 31. Small patch to Sieve code - explicitly set From: when generating an
3244 32. Exim crashed if a remote delivery caused a very long error message to be
3245 recorded - for instance if somebody sent an entire SpamAssassin report back
3246 as a large number of 550 error lines. This bug was coincidentally fixed by
3247 increasing the size of one of Exim's internal buffers (big_buffer) that
3248 happened as part of the Exiscan merge. However, to be on the safe side, I
3249 have made the code more robust (and fixed the comments that describe what
3252 33. Some experimental protocols are using DNS PTR records for new purposes. The
3253 keys for these records are domain names, not reversed IP addresses. The
3254 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3255 leaves it alone. Component reversal etc. now happens only for IP addresses.
3258 34. The host_aton() function is supposed to be passed a string that is known
3259 to be a valid IP address. However, in the case of IPv6 addresses, it was
3260 not checking this. This is a hostage to fortune. Exim now panics and dies
3261 if the condition is not met. A case was found where this could be provoked
3262 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3263 components; fortuitously, this particular loophole had already been fixed
3264 by change 4.50/55 or 4.44/33 above.
3266 If there are any other similar loopholes, the new check in host_aton()
3267 itself should stop them being exploited. The report I received stated that
3268 data on the command line could provoke the exploit when Exim was running as
3269 exim, but did not say which command line option was involved. All I could
3270 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3271 running as the user.
3274 35. There was a buffer overflow vulnerability in the SPA authentication code
3275 (which came originally from the Samba project). I have added a test to the
3276 spa_base64_to_bits() function which I hope fixes it.
3279 36. The daemon start-up calls getloadavg() while still root for those OS that
3280 need the first call to be done as root, but it missed one case: when
3281 deliver_queue_load_max is set with deliver_drop_privilege. This is
3282 necessary for the benefit of the queue runner, because there is no re-exec
3283 when deliver_drop_privilege is set.
3285 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3286 was in use, was not putting the data itself into the right store pool;
3287 consequently, it could be overwritten for a subsequent message in the same
3288 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3291 38. Sometimes the final signoff response after QUIT could fail to get
3292 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3293 < 0 before doing a fflush(). This bug looks as though it goes back to the
3294 introduction of TLS in release 3.20, but "sometimes" must have been rare
3295 because the tests only now provoked it.
3301 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3302 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3303 Now it returns an integer. A number of calls were still expecting a T/F
3304 return. Fortuitously, in all cases, the tests worked in OK situations,
3305 which is the norm. However, things would have gone wrong on any write
3306 failures on the smtp file descriptor. This function is used when sending
3307 messages over SMTP and also when doing verify callouts.
3309 2. When Exim is called to do synchronous delivery of a locally submitted
3310 message (the -odf or -odi options), it no longer closes stderr before doing
3313 3. Implemented the mua_wrapper option.
3315 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3317 5. Implemented the functions header_remove(), header_testname(),
3318 header_add_at_position(), and receive_remove_recipient(), and exported them
3321 6. If an ACL "warn" statement specified the addition of headers, Exim already
3322 inserted X-ACL-Warn: at the start if there was no header name. However, it
3323 was not making this test for the second and subsequent header lines if
3324 there were newlines in the string. This meant that an invalid header could
3325 be inserted if Exim was badly configured.
3327 7. Allow an ACL "warn" statement to add header lines at the start or after all
3328 the Received: headers, as well as at the end.
3330 8. Added the rcpt_4xx retry error code.
3332 9. Added postmaster_mailfrom=xxx to callout verification option.
3334 10. Added mailfrom=xxxx to the callout verification option, for verify=
3337 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3338 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3340 12. Inserted some casts to stop certain compilers warning when using pointer
3341 differences as field lengths or precisions in printf-type calls (mostly
3342 affecting debugging statements).
3344 13. Added optional readline() support for -be (dynamically loaded).
3346 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3347 same clock tick as a message's arrival, so that its received time was the
3348 same as the "first fail" time on the retry record, and that message
3349 remained on the queue past the ultimate address timeout, every queue runner
3350 would try a delivery (because it was past the ultimate address timeout) but
3351 after another failure, the ultimate address timeout, which should have then
3352 bounced the address, did not kick in. This was a "< instead of <=" error;
3353 in most cases the first failure would have been in the next clock tick
3354 after the received time, and all would be well.
3356 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3357 being recognized when the domain list was tested by the match_domain
3358 condition in an expansion string.
3360 16. Added the ${str2b64: operator.
3362 17. Exim was always calling setrlimit() to set a large limit for the number of
3363 processes, without checking whether the existing limit was already
3364 adequate. (It did check for the limit on file descriptors.) Furthermore,
3365 errors from getrlimit() and setrlimit() were being ignored. Now they are
3366 logged to the main and panic logs, but Exim does carry on, to try to do its
3367 job under whatever limits there are.
3369 18. Imported PCRE 5.0.
3371 19. Trivial typo in log message " temporarily refused connection" (the leading
3374 20. If the log selector return_path_on_delivery was set and an address was
3375 redirected to /dev/null, the delivery process crashed because it assumed
3376 that a return path would always be set for a "successful" delivery. In this
3377 case, the whole delivery is bypassed as an optimization, and therefore no
3380 21. Internal re-arrangement: the function for sending a challenge and reading
3381 a response while authentication was assuming a zero-terminated challenge
3382 string. It's now changed to take a pointer and a length, to allow for
3383 binary data in such strings.
3385 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3387 23. Exim was not respecting finduser_retries when seeking the login of the
3388 uid under which it was called; it was always trying 10 times. (The default
3389 setting of finduser_retries is zero.) Also, it was sleeping after the final
3390 failure, which is pointless.
3392 24. Implemented tls_on_connect_ports.
3394 25. Implemented acl_smtp_predata.
3396 26. If the domain in control=submission is set empty, Exim assumes that the
3397 authenticated id is a complete email address when it generates From: or
3398 Sender: header lines.
3400 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3401 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3402 chown and chgrp in /bin and hostname in /usr/bin.
3404 28. Exim was keeping the "process log" file open after each use, just as it
3405 does for the main log. This opens the possibility of it remaining open for
3406 long periods when the USR1 signal hits a daemon. Occasional processlog
3407 errors were reported, that could have been caused by this. Anyway, it seems
3408 much more sensible not to leave this file open at all, so that is what now
3411 29. The long-running daemon process does not normally write to the log once it
3412 has entered its main loop, and it closes the log before doing so. This is
3413 so that log files can straightforwardly be renamed and moved. However,
3414 there are a couple of unusual error situations where the daemon does write
3415 log entries, and I had neglected to close the log afterwards.
3417 30. The text of an SMTP error response that was received during a remote
3418 delivery was being truncated at 512 bytes. This is too short for some of
3419 the long messages that one sometimes sees. I've increased the limit to
3422 31. It is now possible to make retry rules that apply only when a message has a
3423 specific sender, in particular, an empty sender.
3425 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3426 it possible to be selective about when SMTP synchronization is enforced.
3428 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3430 32. Implemented hosts_connection_nolog.
3432 33. Added an ACL for QUIT.
3434 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3437 35. Added mailbox_size and mailbox_filecount to appendfile.
3439 36. Added control = no_multiline_responses to ACLs.
3441 37. There was a bug in the logic of the code that waits for the clock to tick
3442 in the case where the clock went backwards by a substantial amount such
3443 that the microsecond fraction of "now" was more than the microsecond
3444 fraction of "then" (but the whole seconds number was less).
3446 38. Added support for the libradius Radius client library this is found on
3447 FreeBSD (previously only the radiusclient library was supported).
3453 1. When certain lookups returned multiple values in the form name=value, the
3454 quoting of the values was not always being done properly. Specifically:
3455 (a) If the value started with a double quote, but contained no whitespace,
3457 (b) If the value contained whitespace other than a space character (i.e.
3458 tabs or newlines or carriage returns) it was not quoted.
3459 This fix has been applied to the mysql and pgsql lookups by writing a
3460 separate quoting function and calling it from the lookup code. The fix
3461 should probably also be applied to nisplus, ibase and oracle lookups, but
3462 since I cannot test any of those, I have not disturbed their existing code.
3464 2. A hit in the callout cache for a specific address caused a log line with no
3465 reason for rejecting RCPT. Now it says "Previous (cached) callout
3466 verification failure".
3468 3. There was an off-by-one bug in the queryprogram router. An over-long
3469 return line was truncated at 256 instead of 255 characters, thereby
3470 overflowing its buffer with the terminating zero. As well as fixing this, I
3471 have increased the buffer size to 1024 (and made a note to document this).
3473 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3474 when Exim is waiting for an SMTP response from a remote server, Exim
3475 restarts its select() call on the socket, thereby resetting its timeout.
3476 This is not a problem when such interrupts are rare. Somebody set up a cron
3477 job to run exiwhat every 2 minutes, which is less than the normal select()
3478 timeout (5 or 10 minutes). This meant that the select() timeout never
3479 kicked in because it was always reset. I have fixed this by comparing the
3480 time when an interrupt arrives with the time at the start of the first call
3481 to select(). If more time than the timeout has elapsed, the interrupt is
3482 treated as a timeout.
3484 5. Some internal re-factoring in preparation for the addition of Sieve
3485 extensions (by MH). In particular, the "personal" test is moved to a
3486 separate function, and given an option for scanning Cc: and Bcc: (which is
3487 not set for Exim filters).
3489 6. When Exim created an email address using the login of the caller as the
3490 local part (e.g. when creating a From: or Sender: header line), it was not
3491 quoting the local part when it contained special characters such as @.
3493 7. Installed new OpenBSD configuration files.
3495 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3496 try to make them clearer.
3498 9. Callout options, other than the timeout value, were being ignored when
3499 verifying sender addresses in header lines. For example, when using
3501 verify = header_sender/callout=no_cache
3503 the cache was (incorrectly) being used.
3505 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3506 only the Cygwin environment.
3508 11. When return_path_on_delivery was set as a log selector, if different remote
3509 addresses in the same message used different return paths and parallel
3510 remote delivery occurred, the wrong values would sometimes be logged.
3511 (Whenever a remote delivery process finished, the return path value from
3512 the most recently started remote delivery process was logged.)
3514 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3515 lines when AUTH and/or TLS are in use. This is the "received protocol"
3516 field. Exim used to use "asmtp" for authenticated SMTP, without any
3517 indication (in the protocol name) for TLS use. Now it follows the RFC and
3518 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3519 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3520 names appear in log lines as well as in Received: header lines.
3522 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3523 extensions, and comparison tests, and to fix some bugs.
3525 14. Changes to the "personal" filter test:
3527 (1) The test was buggy in that it was just doing the equivalent of
3528 "contains" tests on header lines. For example, if a user's address was
3529 anne@some.where, the "personal" test would incorrectly be true for
3531 To: susanne@some.where
3533 This test is now done by extracting each address from the header in turn,
3534 and checking the entire address. Other tests that are part of "personal"
3535 are now done using regular expressions (for example, to check local parts
3536 of addresses in From: header lines).
3538 (2) The list of non-personal local parts in From: addresses has been
3539 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3540 taken from the Sieve specification recommendations.
3542 (3) If the message contains any header line starting with "List-" it is
3543 treated as non-personal.
3545 (4) The test for "circular" in the Subject: header line has been removed
3546 because it now seems ill-conceived.
3548 15. Minor typos in src/EDITME comments corrected.
3550 16. Installed latest exipick from John Jetmore.
3552 17. If headers_add on a router specified a text string that was too long for
3553 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3554 of string_sprintf() is now avoided.
3556 18. $message_body_size was not set (it was always zero) when running the DATA
3557 ACL and the local_scan() function.
3559 19. For the "mail" command in an Exim filter, no default was being set for
3560 the once_repeat time, causing a random time value to be used if "once" was
3561 specified. (If the value happened to be <= 0, no repeat happened.) The
3562 default is now 0s, meaning "never repeat". The "vacation" command was OK
3563 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3564 (I found it when inspecting the code).
3566 20. There is now an overall timeout for performing a callout verification. It
3567 defaults to 4 times the callout timeout, which applies to individual SMTP
3568 commands during the callout. The overall timeout applies when there is more
3569 than one host that can be tried. The timeout is checked before trying the
3570 next host. This prevents very long delays if there are a large number of
3571 hosts and all are timing out (e.g. when the network connections are timing
3572 out). The value of the overall timeout can be changed by specifying an
3573 additional sub-option for "callout", called "maxwait". For example:
3575 verify = sender/callout=5s,maxwait=20s
3577 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3578 to the end before writing, but this should make it even safer).
3580 22. Exim was forgetting that it had advertised PIPELINING for the second and
3581 subsequent messages on an SMTP connection. It was also not resetting its
3582 memory on STARTTLS and an internal HELO.
3584 23. When Exim logs an SMTP synchronization error within a session, it now
3585 records whether PIPELINING has been advertised or not.
3587 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3588 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3589 rather than long int.
3591 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3593 26. Added the never_mail option to autoreply.
3599 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3600 crash if the getsockname() call failed; this can happen if a connection is
3601 closed very soon after it is established. The problem was simply in the
3602 order in which certain operations were done, causing Exim to try to write
3603 to the SMTP stream before it had set up the file descriptor. The bug has
3604 been fixed by making things happen in the correct order.
3610 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3611 before the connection was closed, thus losing the rejection response.
3613 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3614 some early Solaris releases, but causes trouble in current releases where
3615 socklen_t is defined.
3617 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3620 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3623 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3624 that releasing the top of it at the end releases what was used for sub-
3625 expansions (unless the block got too big). However, discard this block if
3626 the first thing is a variable or header, so that we can use its block when
3627 it is dynamic (useful for very large $message_headers, for example).
3629 6. Lookups now cache *every* query, not just the most recent. A new, separate
3630 store pool is used for this. It can be recovered when all lookup caches are
3631 flushed. Lookups now release memory at the end of their result strings.
3632 This has involved some general refactoring of the lookup sources.
3634 7. Some code has been added to the store_xxx() functions to reduce the amount
3635 of flapping under certain conditions.
3637 8. log_incoming_interface used to affect only the <= reception log lines. Now
3638 it causes the local interface and port to be added to several more SMTP log
3639 lines, for example "SMTP connection from", and rejection lines.
3641 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3643 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3645 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3646 could be overwritten at the end of the current message (or the start of a
3647 new message if it was set in a HELO ACL). The value is now preserved for
3648 the duration of the SMTP connection.
3650 12. If a transport had a headers_rewrite setting, and a matching header line
3651 contained an unqualified address, that address was qualified, even if it
3652 did not match any rewriting rules. The underlying bug was that the values
3653 of the flags that permit the existence of unqualified sender and recipient
3654 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3655 non-local messages, and by -bnq for local messages) were not being
3656 preserved with the message after it was received.
3658 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3659 "next input=" as part of the text comprising the host identity instead of
3660 the correct text. The code was using the same buffer for two different
3661 strings. However, depending on which order the printing function evaluated
3662 its arguments, the bug did not always show up. Under Linux, for example, my
3663 test suite worked just fine.
3665 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3666 doesn't work with some older versions of Perl. It has been changed to "my",
3667 which in any case is probably the better facility to use.
3669 15. A really picky compiler found some instances of statements for creating
3670 error messages that either had too many or two few arguments for the format
3673 16. The size of the buffer for calls to the DNS resolver has been increased
3674 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3675 for addresses that have a lot of PTR records. This alleviates a problem; it
3676 does not fully solve it.
3678 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3679 buffer now truncates the list and logs the incident, which is the same
3680 action as happens when Exim is looking up a host name and its aliases.
3681 Previously in this situation something unpredictable would happen;
3682 sometimes it was "internal error: store_reset failed".
3684 18. If a server dropped the connection unexpectedly when an Exim client was
3685 using GnuTLS and trying to read a response, the client delivery process
3686 crashed while trying to generate an error log message.
3688 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3689 string, for example:
3691 warn message = H1: something\nH2: something
3693 the text was added as a single header line from Exim's point of view
3694 though it ended up OK in the delivered message. However, searching for the
3695 second and subsequent header lines using $h_h2: did not work. This has been
3696 fixed. Similarly, if a system filter added multiple headers in this way,
3697 the routers could not see them.
3699 20. Expanded the error message when iplsearch is called with an invalid key to
3700 suggest using net-iplsearch in a host list.
3702 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3703 ACLs are no longer actually imposed (and a message to that effect is
3706 22. If a "gecos" field in a passwd entry contained escaped characters, in
3707 particular, if it contained a \" sequence, Exim got it wrong when building
3708 a From: or a Sender: header from that name. A second bug also caused
3709 incorrect handling when an unquoted " was present following a character
3710 that needed quoting.
3712 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3713 was not being matched caselessly.
3715 24. Arranged for all hyphens in the exim.8 source to be escaped with
3718 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3719 properly. Recipient callout cache records were still being keyed to include
3720 the sender, even when use_sender was set false. This led to far more
3721 callouts that were necessary. The sender is no longer included in the key
3722 when use_sender is false.
3724 26. Added "control = submission" modifier to ACLs.
3726 27. Added the ${base62d: operator to decode base 62 numbers.
3728 28. dnsdb lookups can now access SRV records.
3730 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3731 the configuration file.
3733 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3734 (-v) mode. This makes the output for a verbose queue run more intelligible.
3736 31. Added a use_postmaster feature to recipient callouts.
3738 32. Added the $body_zerocount variable, containing the number of binary zero
3739 bytes in the message body.
3741 33. The time of last modification of the "new" subdirectory is now used as the
3742 "mailbox time last read" when there is a quota error for a maildir
3745 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3747 35. Added +ignore_unknown as a special item in host lists.
3749 36. Code for decoding IPv6 addresses in host lists is now included, even if
3750 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3751 address was recognized as an IP address, but was then not correctly decoded
3752 into binary, causing unexpected and incorrect effects when compared with
3759 1. Very minor rewording of debugging text in manualroute to say "list of
3760 hosts" instead of "hostlist".
3762 2. If verify=header_syntax was set, and a header line with an unqualified
3763 address (no domain) and a large number of spaces between the end of the
3764 name and the colon was received, the reception process suffered a buffer
3765 overflow, and (when I tested it) crashed. This was caused by some obsolete
3766 code that should have been removed. The fix is to remove it!
3768 3. When running in the test harness, delay a bit after writing a bounce
3769 message to get a bit more predictability in the log output.
3771 4. Added a call to search_tidyup() just before forking a reception process. In
3772 theory, someone could use a lookup in the expansion of smtp_accept_max_
3773 per_host which, without the tidyup, could leave open a database connection.
3775 5. Added the variables $recipient_data and $sender_data which get set from a
3776 lookup success in an ACL "recipients" or "senders" condition, or a router
3777 "senders" option, similar to $domain_data and $local_part_data.
3779 6. Moved the writing of debug_print from before to after the "senders" test
3782 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3783 problems for message scanning, either using a data ACL, or using
3784 local_scan() because the Received: header was not generated till after they
3785 were called (in order to set the time as the time of reception completion).
3786 I have revised the way this works. The header is now generated after the
3787 body is received, but before the ACL or local_scan() are called. After they
3788 are run, the timestamp in the header is updated.
3794 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3795 before starting a queue runner without re-exec. This happened only when
3796 deliver_drop_privilege was set or when the Exim user was set to root. The
3797 effect of the bug was that timeouts during subsequent deliveries caused
3798 crashes instead of being properly handled. The handler is now left at its
3799 default (and expected) setting.
3801 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3802 message, again when deliver_drop_privilege is set or Exim is run as root.
3803 The bug described in (1) was not present in this case, but the tidying up
3804 of the other signals was missing. I have made the two cases consistent.
3806 3. The ignore_target_hosts setting on a manualroute router was being ignored
3807 for hosts that were looked up using the /MX notation.
3809 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3812 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3813 operated on the sender address. After changing the $sender_address to <>
3814 for the sender address verify, Exim was re-instated it as the original
3815 (before rewriting) address, but remembering that it had rewritten it, so it
3816 wasn't rewriting it again. This bug also had the effect of breaking the
3817 sender address verification caching when the sender address was rewritten.
3819 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3820 This has been changed so that if the ip literal address matches
3821 ignore_target_hosts, the router declines.
3823 7. Added expansion conditions match_domain, match_address, and match_local_
3824 part (NOT match_host).
3826 8. The placeholder for the Received: header didn't have a length field set.
3828 9. Added code to Exim itself and to exim_lock to test for a specific race
3829 condition that could lead to file corruption when using MBX delivery. The
3830 issue is with the lockfile that is created in /tmp. If this file is removed
3831 after a process has opened it but before that process has acquired a lock,
3832 there is the potential for a second process to recreate the file and also
3833 acquire a lock. This could lead to two Exim processes writing to the file
3834 at the same time. The added code performs the same test as UW imapd; it
3835 checks after acquiring the lock that its file descriptor still refers to
3836 the same named file.
3838 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3839 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3842 11. Added the smtp_active_hostname option. If used, this will typically be made
3843 to depend on the incoming interface address. Because $interface_address is
3844 not set up until the daemon has forked a reception process, error responses
3845 that can happen earlier (such as "too many connections") no longer contain
3848 12. If an expansion in a condition on a "warn" statement fails because a lookup
3849 defers, the "warn" statement is abandoned, and the next ACL statement is
3850 processed. Previously this caused the whole ACL to be aborted.
3852 13. Added the iplsearch lookup type.
3854 14. Added ident_timeout as a log selector.
3856 15. Added tls_certificate_verified as a log selector.
3858 16. Added a global option tls_require_ciphers (compare the smtp transport
3859 option of the same name). This controls incoming TLS connections.
3861 17. I finally figured out how to make tls_require_ciphers do a similar thing
3862 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3863 before starting the TLS session.
3865 18. Tabs are now shown as \t in -bP output.
3867 19. If the log selector return_path_on_delivery was set, Exim crashed when
3868 bouncing a message because it had too many Received: header lines.
3870 20. If two routers both had headers_remove settings, and the first one included
3871 a superfluous trailing colon, the final name in the first list and the
3872 first name in the second list were incorrectly joined into one item (with a
3873 colon in the middle).
3879 1. Added -C and -D options to the exinext utility, mainly to make it easier
3880 to include in the automated testing, but these could be helpful when
3881 multiple configurations are in use.
3883 2. The exinext utility was not formatting the output nicely when there was
3884 an alternate port involved in the retry record key, nor when there was a
3885 message id as well (for retries that were specific to a specific message
3886 and a specific host). It was also confused by IPv6 addresses, because of
3887 the additional colons they contain. I have fixed the IPv4 problem, and
3888 patched it up to do a reasonable job for IPv6.
3890 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3891 delivery, the log line now contains "pipelined" if PIPELINING was used.
3893 4. An SMTP transport process used to panic and die if the bind() call to set
3894 an explicit outgoing interface failed. This has been changed; it is now
3895 treated in the same way as a connect() failure.
3897 5. A reference to $sender_host_name in the part of a conditional expansion
3898 that was being skipped was still causing a DNS lookup. This no longer
3901 6. The def: expansion condition was not recognizing references to header lines
3902 that used bh_ and bheader_.
3904 7. Added the _cache feature to named lists.
3906 8. The code for checking quota_filecount in the appendfile transport was
3907 allowing one more file than it should have been.
3909 9. For compatibility with Sendmail, the command line option
3917 and sets the incoming protocol and host name (for trusted callers). The
3918 host name and its colon can be omitted when only the protocol is to be set.
3919 Note the Exim already has two private options, -pd and -ps, that refer to
3920 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3921 "s", but I don't think that's a major issue.
3923 10. A number of refactoring changes to the code, none of which should affect
3926 (a) The number of logging options was getting close to filling up the
3927 32-bit word that was used as a bit map. I have split them into two classes:
3928 those that are passed in the argument to log_write(), and those that are
3929 only ever tested independently outside of that function. These are now in
3930 separate 32-bit words, so there is plenty of room for expansion again.
3931 There is no change in the user interface or the logging behaviour.
3933 (b) When building, for example, log lines, the code previously used a
3934 macro that called string_cat() twice, in order to add two strings. This is
3935 not really sufficiently general. Furthermore, there was one instance where
3936 it was actually wrong because one of the argument was used twice, and in
3937 one call a function was used. (As it happened, calling the function twice
3938 did not affect the overall behaviour.) The macro has been replaced by a
3939 function that can join an arbitrary number of extra strings onto a growing
3942 (c) The code for expansion conditions now uses a table and a binary chop
3943 instead of a serial search (which was left over from when there were very
3944 few conditions). Also, it now recognizes conditions like "pam" even when
3945 the relevant support is not compiled in: a suitably worded error message is
3946 given if an attempt is made to use such a condition.
3948 11. Added ${time_interval:xxxxx}.
3950 12. A bug was causing one of the ddress fields not to be passed back correctly
3951 from remote delivery subprocesses. The field in question was not being
3952 subsequently used, so this caused to problems in practice.
3954 13. Added new log selectors queue_time and deliver_time.
3956 14. Might have fixed a bug in maildirsizefile handling that threw up
3957 "unexpected character" debug warnings, and recalculated the data
3958 unnecessarily. In any case, I expanded the warning message to give more
3961 15. Added the message "Restricted characters in address" to the statements in
3962 the default ACL that block characters like @ and % in local parts.
3964 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3965 Three changes have been made:
3967 (a) There was a serious bug; a negative response to MAIL caused the whole
3968 recipient domain to be cached as invalid, thereby blocking all messages
3969 to all local parts at the same domain, from all senders. This bug has
3970 been fixed. The domain is no longer cached after a negative response to
3971 MAIL if the sender used is not empty.
3973 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3976 (c) A new callout option, "use_sender" has been added for people who want
3977 the modified behaviour.
3983 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3986 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3987 indeed breaks things for older releases.
3989 3. Added additional logging to the case where there is a problem reading data
3990 from a filter that is running in a subprocess using a pipe, in order to
3991 try to track down a specific problem.
3993 4. Testing facility fudge: when running in the test harness and attempting
3994 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3995 getting "No route to host". Convert this to a timeout.
3997 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4000 6. Some OS don't have socklen_t but use size_t instead. This affects the
4001 fifth argument of getsockopt() amongst other things. This is now
4002 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4003 can be set for individual OS. I have set it for SunOS5, OSF1, and
4004 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4005 some earlier ones do not.
4007 7. Change 4.30/15 was not doing the test caselessly.
4009 8. The standard form for an IPv6 address literal was being rejected by address
4010 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4011 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4012 this, as well as the form without the "IPv6" on the front (but only when
4013 address literals are enabled, of course).
4015 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4017 10. Exim crashed if a message with an empty sender address specified by -f
4018 encountered a router with an errors_to setting. This could be provoked only
4019 by a command such as
4023 where an empty string was supplied; "<>" did not hit this bug.
4025 11. Installed PCRE release 4.5.
4027 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4028 remained set. It is now erased.
4030 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4031 times from message ids (which are base 36 rather than the normal 62).
4033 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4034 were being counted as actual protocol errors, and logged if the log
4035 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4036 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4037 and DATA following a set of rejected RCPTs do not count as protocol errors.
4038 In other words, Exim assumes they were pipelined, though this may not
4039 actually be the case. Of course, in all cases the client gets an
4040 appropriate error code.
4042 15. If a lookup fails in an ACL condition, a message about the failure may
4043 be available; it is used if testing the ACL cannot continue, because most
4044 such messages specify what the cause of the deferral is. However, some
4045 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4046 that caused an old message to be retained and used if a later statement
4047 caused a defer, replacing the real cause of the deferral.
4049 16. If an IP address had so many PTR records that the DNS lookup buffer
4050 was not large enough to hold them, Exim could crash while trying to process
4051 the truncated data. It now detects and logs this case.
4053 17. Further to 4.21/58, another change has been made: if (and only if) the
4054 first line of a message (the first header line) ends with CRLF, a bare LF
4055 in a subsequent header line has a space inserted after it, so as not to
4056 terminate the header.
4058 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4059 unnecessarily, used atoi() instead of strtol(), and didn't check the
4060 termination when getting file sizes from file names by regex.
4062 19. Completely re-implemented the support for maildirsize files, in the light
4063 of a number of problems with the previous contributed implementation
4064 (4.30/29). In particular:
4066 . If the quota is zero, the maildirsize file is maintained, but no quota is
4069 . If the maildir directory does not exist, it is created before any attempt
4070 to write a maildirsize file.
4072 . The quota value in the file is just a cache; if the quota is changed in
4073 the transport, the new value overrides.
4075 . A regular expression is available for excluding directories from the
4078 20. The autoreply transport checks the characters in options that define the
4079 message's headers; it allows continued headers, but it was checking with
4080 isspace() after an embedded newline instead of explicitly looking for a
4083 21. If all the "regular" hosts to which an address was routed had passed their
4084 expiry times, and had not reached their retry times, the address was
4085 bounced, even if fallback hosts were defined. Now Exim should go on to try
4088 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4089 equivalent code in the SMTP transport. Some hosts send humungous responses
4090 to HELO/EHLO, more than 1024 it seems.
4092 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4093 gives compiler warnings in some environments. I've now done it "properly",
4096 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4097 (because of problems with the built-in one) was declared to return uschar *
4098 instead of char *, causing compiler failure.
4100 25. Fixed a file descriptor leak when processing alias/forward files.
4102 26. Fixed a minor format string issue in dbfn.c.
4104 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4106 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4107 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4110 29. When a sender address is verified, it is cached, to save repeating the test
4111 when there is more than one recipient in a message. However, when the
4112 verification involves a callout, it is possible for different callout
4113 options to be set for different recipients. It is too complicated to keep
4114 track of this in the cache, so now Exim always runs a verification when a
4115 callout is required, relying on the callout cache for the optimization.
4116 The overhead is duplication of the address routing, but this should not be
4119 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4120 to be verified with callout=postmaster, and the main callout worked but the
4121 postmaster check failed, the verification correctly failed. However, if a
4122 subsequent RCPT command asked for sender verification *without* the
4123 postmaster check, incorrect caching caused this verification also to fail,
4126 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4127 it was not caching the DNS options (qualify_single, search_parents) that
4128 were used when the lookup failed. A subsequent lookup with different
4129 options therefore always gave the same answer, though there were cases
4130 where it should not have. (Example: a "domains = !$mx_any" option on a
4131 dnslookup router: the "domains" option is always processed without any
4132 widening, but the router might have qualify_single set.) Now Exim uses the
4133 cached value only when the same options are set.
4135 32. Added John Jetmore's "exipick" utility to the distribution.
4137 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4138 than a timeout (e.g. a certificate is required, and is not provided), an
4139 Exim server now closes the connection immediately. Previously it waited for
4140 the client to close - but if the client is SSL, it seems that they each
4141 wait for each other, leading to a delay before one of them times out.
4143 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4144 maintained 0.8.x compatibility because I don't think many are using it, and
4145 it is clearly obsolete.
4147 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4150 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4151 client certificate was expired. A simple patch fixes this, though I don't
4152 understand the full logic of why the verify callback is called multiple
4155 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4156 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4157 which causes problems with some clients (such as the Certicom SSL Plus
4158 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4159 disables the coutermeasure allowing Eudora to connect."
4161 38. Exim was not checking that a write() to a log file succeeded. This could
4162 lead to Bad Things if a log got too big, in particular if it hit a file
4163 size limit. Exim now panics and dies if it cannot write to a log file, just
4164 as it does if it cannot open a log file.
4166 39. Modified OS/Makefile-Linux so that it now contains
4168 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4170 The two -D definitions ensure that Exim is compiled with large file
4171 support, which makes it possible to handle log files that are bigger than
4174 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4175 instance) a domain was checked against a named list that involved a lookup,
4176 causing $domain_data to be set, then another domain was checked against the
4177 same list, then the first domain was re-checked, the value of $domain_data
4178 after the final check could be wrong. In particular, if the second check
4179 failed, it could be set empty. This bug probably also applied to
4182 41. The strip_trailing_dot option was not being applied to the address given
4183 with the -f command-line option.
4185 42. The code for reading a message's header from the spool was incrementing
4186 $received_count, but never initializing it. This meant that the value was
4187 incorrect (doubled) while delivering a message in the same process in which
4188 it was received. In the most common configuration of Exim, this never
4189 happens - a fresh exec is done - but it can happen when
4190 deliver_drop_privilege is set.
4192 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4193 it now includes up to 150 characters of the unexpected data in the log
4196 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4197 and building data strings. The size of both of these buffers was 10 000
4198 bytes - far larger than anybody would *ever* want, thought I. Needless to
4199 say, somebody hit the limit. I have increased the maximum line length to
4200 20 000 and the maximum data length of concatenated lines to 100 000. I have
4201 also fixed two bugs, because there was no checking on these buffers. Tsk,
4202 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4203 buffer is too small.
4205 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4206 lsearch lookups. Now it does.
4208 46. When parsing a route_list item in a manualroute router, a fixed-length
4209 buffer was used for the list of hosts. I made this 1024 bytes long,
4210 thinking that nobody would ever have a list of hosts that long. Wrong.
4211 Somebody had a whole pile of complicated expansion conditions, and the
4212 string was silently truncated, leading to an expansion error. It turns out
4213 that it is easier to change to an unlimited length (owing to other changes
4214 that have happened since this code was originally written) than to build
4215 structure for giving a limitation error. The length of the item that
4216 expands into the list of hosts is now unlimited.
4218 47. The lsearch lookup could not handle data where the length of text line was
4219 more than 4095 characters. Such lines were truncated, leading to shortened
4220 data being returned. It should now handle lines of any length.
4222 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4223 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4226 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4227 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4228 no longer happens in comment lines. A list of replacements is now placed
4229 at the head of all of the source files, except those whose only change is
4230 to replace PERL_COMMAND in the very first #! line.
4232 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4233 messages on the queue, with a bottom-up merge sort, using code contributed
4234 by Michael Haardt. This should make operations like -bp somewhat faster on
4235 large queues. It won't affect queue runners, except when queue_run_in_order
4238 51. Installed eximstats 1.31 in the distribution.
4240 52. Added support for SRV lookups to the dnslookup router.
4242 53. If an ACL referred to $message_body or $message_body_end, the value was not
4243 reset for any messages that followed in the same SMTP session.
4245 54. The store-handling optimization for building very long strings was not
4246 differentiating between the different store pools. I don't think this
4247 actually made any difference in practice, but I've tidied it.
4249 55. While running the routers to verify a sender address, $sender_address
4250 was still set to the sender address. This is wrong, because when routing to
4251 send a bounce to the sender, it would be empty. Therefore, I have changed
4252 it so that, while verifying a sender address, $sender_address is set to <>.
4253 (There is no change to what happens when verifying a recipient address.)
4255 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4256 target A or AAAA records (if not already returned) without resetting the
4257 qualify_single or search_parents options of the DNS resolver. These are
4258 inappropriate in this case because the targets of MX and SRV records must
4259 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4260 target that, when qualified, matched something in the local domain. These
4261 two options are now turned off when doing these lookups.
4263 57. It seems that at least some releases of Reiserfs (which does not have the
4264 concept of a fixed number of inodes) returns zero and not -1 for the
4265 number of available inodes. This interacted badly with check_spool_inodes,
4266 which assumed that -1 was the "no such thing" setting. What I have done is
4267 to check that the total number of inodes is greater than zero before doing
4268 the test of how many are available.
4270 58. When a "warn" ACL statement has a log_message modifier, the message is
4271 remembered, and not repeated. This is to avoid a lot of repetition when a
4272 message has many recipients that cause the same warning to be written.
4273 Howewer, Exim was preserving the list of already written lines for an
4274 entire SMTP session, which doesn't seem right. The memory is now reset if a
4275 new message is started.
4277 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4278 debugging output unless log_rewrite was also set.
4280 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4281 of (int)(handle) when we know that handle contains (void *)(-1).
4283 61. The Exim daemon panic-logs an error return when it closes the incoming
4284 connection. However "connection reset by peer" seems to be common, and
4285 isn't really an error worthy of noting specially, so that particular error
4288 62. When Exim is trying to find all the local interfaces, it used to panic and
4289 die if the ioctl to get the interface flags failed. However, it seems that
4290 on at least one OS (Solaris 9) it is possible to have an interface that is
4291 included in the list of interfaces, but for which you get a failure error
4292 for this call. This happens when the interface is not "plumbed" into a
4293 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4294 failure of the "get flags" call assumes that the interface is down.
4296 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4297 makes life easier for people who are doing arithmetic on fields extracted
4298 from dates, where you often get leading zeros that should not be
4299 interpreted as octal.
4301 64. Added qualify_domain to the redirect router, to override the global
4304 65. If a pathologically long header line contained very many addresses (the
4305 report of this problem mentioned 10 000) and each of them was rewritten,
4306 Exim could use up a very large amount of memory. (It kept on making new
4307 copies of the header line as it rewrote, and never released the old ones.)
4308 At the expense of a bit more processing, the header rewriting function has
4309 been changed so that it no longer eats memory in this way.
4311 66. The generation of the Received: header has been moved from the time that a
4312 message starts to be received, to the time that it finishes. The timestamp
4313 in the Received: header should now be very close to that of the <= log
4314 line. There are two side-effects of this change:
4316 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4317 logged header lines no longer include the local Received: line, because
4318 it has not yet been created. The same applies to a copy of the message
4319 that is returned to a non-SMTP sender when a message is rejected.
4321 (b) When a filter file is tested using -bf, no additional Received: header
4322 is added to the test message. After some thought, I decided that this
4325 This change does not affect the value of $received_for. It is still set
4326 after address rewriting, but before local_scan() is called.
4328 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4330 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4331 gave an unhelpful panic error message, and a defer error. I have managed to
4332 change this behaviour so that it now rejects any supplied certificate,
4333 which seems right, as the list of acceptable certificates is empty.
4335 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4336 gave an unhelpful defer error. I have not managed to make this reject any
4337 supplied certificates, but the error message it gives is "no certificate
4338 supplied", which is not helpful.
4340 70. exigrep's output now also includes lines that are not associated with any
4341 message, but which match the given pattern. Implemented by a patch from
4342 Martin Sluka, which also tidied up the Perl a bit.
4344 71. Recipient callout verification, like sender verification, was using <> in
4345 the MAIL FROM command. This isn't really the right thing, since the actual
4346 sender may affect whether the remote host accepts the recipient or not. I
4347 have changed it to use the actual sender in the callout; this means that
4348 the cache record is now keyed on a recipient/sender pair, not just the
4349 recipient address. There doesn't seem to be a real danger of callout loops,
4350 since a callout by the remote host to check the sender would use <>.
4351 [SEE ABOVE: changed after hitting problems.]
4353 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4354 temporary errors. However, in the case of such a code being given after
4355 the end of a data transmission (i.e. after ".") Exim was failing to write
4356 a retry record for the message. (Yes, there was some broken host that was
4357 actually sending 8xx at this point.)
4359 73. An unknown lookup type in a host list could cause Exim to panic-die when
4360 the list was checked. (An example that provoked this was putting <; in the
4361 middle of a list instead of at the start.) If this happened during a DATA
4362 ACL check, a -D file could be left lying around. This kind of configuration
4363 error no longer causes Exim to die; instead it causes a defer errror. The
4364 incident is still logged to the main and panic logs.
4366 74. Buglet left over from Exim 3 conversion. The message "too many messages
4367 in one connection" was written to the rejectlog but not the mainlog, except
4368 when address rewriting (yes!) was being logged.
4370 75. Added write_rejectlog option.
4372 76. When a system filter was run not as root (that is, when system_filter_user
4373 was set), the values of the $n variables were not being returned to the
4374 main process; thus, they were not subsequently available in the $sn
4377 77. Added +return_path_on_delivery log selector.
4379 78. A connection timeout was being treated differently from recipients deferred
4380 when testing hosts_max_try with a message that was older than the host's
4381 retry timeout. (The host should not be counted, thus allowing all hosts to
4382 be tried at least once before bouncing.) This may have been the cause of an
4383 occasionally reported bug whereby a message would remain on the queue
4384 longer than the retry timeout, but would be bounced if a delivery was
4385 forced. I say "may" because I never totally pinned down the problem;
4386 setting up timeout/retry tests is difficult. See also the next item.
4388 79. The ultimate address timeout was not being applied to errors that involved
4389 a combination of host plus message (for example, a timeout on a MAIL
4390 command). When an address resolved to a number of possible hosts, and they
4391 were not all tried for each delivery (e.g. because of hosts_max_try), a
4392 message could remain on the queue longer than the retry timeout.
4394 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4397 81. Fixed an obscure SMTP outgoing bug which required at least the following
4398 conditions: (a) there was another message waiting for the same server;
4399 (b) the server returned 5xx to all RCPT commands in the first message so
4400 that the message was not completed; (c) the server dropped the connection
4401 or gave a negative response to the RSET that Exim sends to abort the
4402 transaction. The observed case was a dropped connection after DATA that had
4403 been sent in pipelining mode. That is, the server had advertised PIPELINING
4404 but was not implementing it correctly. The effect of the bug was incorrect
4405 behaviour, such as trying another host, and this could lead to a crash.
4411 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4412 and daemon.c were passed as pointers to ints; they should have been
4413 pointers to socklen_t variables (which are typically unsigned ints).
4415 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4418 3. Fixed a really odd bug that affected only the testing scheme; patching a
4419 certain fixed string in the binary changed the value of another string that
4420 happened to be identical to the end of the original first string.
4422 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4423 name", it returns that address as the IP address. On some operating
4424 systems (e.g. Solaris), it also passes back the IP address string as the
4425 "host name". However, on others (e.g. Linux), it passes back an empty
4426 string. Exim wasn't checking for this, and was changing the host name to an
4427 empty string, assuming it had been canonicized.
4429 5. Although rare, it is permitted to have more than one PTR record for a given
4430 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4431 all the names associated with an address, because they do in Solaris.
4432 However, it seems that they do not in Linux for data that comes from the
4433 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4434 I found this out when I moved to a new Linux workstation and tried to run
4435 the Exim test suite.
4437 To get round this problem I have changed the code so that it now does its
4438 own call to the DNS to look up PTR records when searching for a host name.
4439 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4440 addresses that are only in /etc/hosts are still found.
4442 This behaviour is, however, controlled by an option called host_lookup_
4443 order, which defaults to "bydns:byaddr". If people want to use the other
4444 order, or indeed, just use one or the other means of lookup, they can
4445 specify it in this variable.
4447 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4448 some operating systems, this comes back from gethostbyaddr() as an empty
4449 string, and this is what Exim used to test for. However, it seems that in
4450 other systems, "." is yielded. Exim now tests for this case too.
4452 7. The values of check_spool_space and check_log_space are now held internally
4453 as a number of kilobytes instead of an absolute number of bytes. If a
4454 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4455 kilobyte. This means that much larger values can be stored.
4457 8. Exim monitor: an attempt to get the action menu when not actually pointing
4458 at a message produces an empty menu entitled "No message selected". This
4459 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4460 no entries in it ("Shell widget menu has zero width and/or height"). So I
4461 have added a single, blank menu entry in this case.
4463 9. Added ${quote_local_part.
4465 10. MIME decoding is now applied to the contents of Subject: header lines when
4468 11. Now that a reference to $sender_host_address automatically causes a reverse
4469 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4470 host lookup before query-style lookups in lists that might use this
4471 variable. This has therefore been abolished, and the "net-" prefix is no
4472 longer necessary for query-style lookups.
4474 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4475 have been a typo for LFLAGS, so it has been changed.
4477 13. The install script calls Exim with "-C /dev/null" in order to find the
4478 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4479 to be output. Howeve, since Exim outputs its version number before the
4480 error, it didn't break the script. It just looked ugly. I fixed this by
4481 always allowing "-C /dev/null" if the caller is root.
4483 14. Ignore overlarge ACL variable number when reading spool file - insurance
4484 against a later release with more variables having written the file.
4486 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4487 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4488 this, as well as the form without the "IPv6" on the front.
4490 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4491 OS/Makefile-Darwin file.
4493 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4494 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4496 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4497 variable. However, some OS use ints for inodes. I've added cast to long int
4498 to get rid of the compiler warning.
4500 19. I had forgotten to lock out "/../" in configuration file names when
4501 ALT_CONFIG_PREFIX was set.
4503 20. Routers used for verification do not need to specify transports. However,
4504 if such a router generated a host list, and callout was configured, Exim
4505 crashed, because it could not find a port number from the (non-existent)
4506 transport. It now assumes port 25 in this circumstance.
4508 21. Added the -t option to exigrep.
4510 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4511 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4512 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4513 them is set, LOOKUP_LSEARCH is forced.
4515 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4517 24. Added sender and host information to the "rejected by local_scan()" log
4518 line; previously there was no indication of these.
4520 25. Added .include_if_exists.
4522 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4523 Linux. It turns out that not all file systems support this. Apparently some
4524 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4525 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4526 not supported on the file descriptor, is now ignored when Exim is trying to
4527 sync a directory. This applies only to Linux.
4529 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4531 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4532 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4535 29. Added support for maildirsize files from supplied patch (modified a bit).
4537 30. The use of :fail: followed by an empty string could lead Exim to respond to
4538 sender verification failures with (e.g.):
4540 550 Verification failed for <xxx>
4541 550 Sender verify failed
4543 where the first response line was missing the '-' that indicates it is not
4544 the final line of the response.
4546 31. The loop for finding the name of the user that called Exim had a hardwired
4547 limit of 10; it now uses the value of finduser_retries, which is used for
4548 all other user lookups.
4550 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4553 33. Exim was neglecting to zero errno before one call of strtol() when
4554 expanding a string and expecting an integer value. On some systems this
4555 resulted in spurious "integer overflow" errors. Also, it was casting the
4556 result into an int without checking.
4558 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4559 did not work. The code looks as if it has *never* worked, though it appears
4560 to have been documented since at least releast 1.62. I have made it work.
4562 35. The "timeout_DNS" error in retry rules, also documented since at least
4563 1.62, also never worked. As it isn't clear exactly what this means, and
4564 clearly it isn't a major issue, I have abolished the feature by treating it
4565 as "timeout", and writing a warning to the main and panic logs.
4567 36. The display of retry rules for -brt wasn't always showing the error code
4570 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4571 timeout_connect_A, timeout_connect_MX.
4573 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4574 to the empty sender.
4576 39. The daemon was not analysing the content of -oX till after it had closed
4577 stderr and disconnected from the controlling terminal. This meant that any
4578 syntax errors were only noted on the panic log, and the return code from
4579 the command was 0. By re-arranging the code a little, I've made the
4580 decoding happen first, so such errors now appear on stderr, and the return
4581 code is 1. However, the actual setting up of the sockets still happens in
4582 the disconnected process, so errors there are still only recorded on the
4585 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4586 connections (as happens on some IP stacks) was logged at start up time as
4587 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4588 it from "IPv6 and IPv4", which means that two separate sockets are being
4591 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4592 says whether AF_INET or AF_INET6 was passed as an argument.
4594 42. Exiwhat output was messed up when time zones were included in log
4597 43. Exiwhat now gives more information about the daemon's listening ports,
4598 and whether -tls-on-connect was used.
4600 44. The "port" option of the smtp transport is now expanded.
4602 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4603 silently ignored. Now an error message is written to the main and panic
4606 46. There's a new ACL modifier called "logwrite" which writes to a log file
4607 as soon as it is encountered.
4609 47. Added $local_user_uid and $local_user_gid at routing time.
4611 48. Exim crashed when trying to verify a sender address that was being
4614 49. Exim was recognizing only a space character after ".include". It now also
4615 recognizes a tab character.
4617 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4618 extracting the relevant information from the specification. The man page no
4619 longer contains scrambled data for the -d option, and I've added a section
4620 at the front about calling Exim under different names.
4622 51. Added "extra_headers" argument to the "mail" command in filter files.
4624 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4627 53. Installed eximstats 1.29.
4629 54. Added transport_filter_timeout as a generic transport option.
4631 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4632 Cc: header lines. This was required by RFC 822, but it not required by RFC
4635 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4636 incoming messages that did not have them. Now it does so only if the
4637 message originates locally, that is, if there is no associated remote host
4638 address. When Resent- header lines are present, this applies to the Resent-
4639 lines rather than the non-Resent- lines.
4641 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4642 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4644 58. Messages for configuration errors now include the name of the main
4645 configuration file - useful now that there may be more than one file in a
4646 list (.included file names were always shown).
4648 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4649 for those rare installations that do not start the daemon as root or run it
4650 setuid root. I've cut out the call to initgroups() if the daemon is not
4653 60. The Exim user and group can now be bound into the binary as text strings
4654 that are looked up at the start of Exim's processing.
4656 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4658 62. Added $mailstore_basename variable.
4660 63. Installed patch to sieve.c from Michael Haardt.
4662 64. When Exim failed to open the panic log after failing to open the main log,
4663 the original message it was trying to log was written to stderr and debug
4664 output, but if they were not available (the usual case in production), it
4665 was lost. Now it is written to syslog before the two lines that record the
4666 failures to open the logs.
4668 65. Users' Exim filters run in subprocesses under the user's uid. It is
4669 possible for a "deliver" command or an alias in a "personal" command to
4670 provoke an address rewrite. If logging of address rewriting is configured,
4671 this fails because the process is not running as root or exim. There may be
4672 a better way of dealing with this, but for the moment (because 4.30 needs
4673 to be released), I have disabled address rewrite logging when running a
4674 filter in a non-root, non-exim process.
4680 1. The buildconfig auxiliary program wasn't quoting the value set for
4681 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4682 not defined. This bug was masked in 4.22 by the effect that was fixed in
4685 2. Some messages that were rejected after a message id was allocated were
4686 shown as "incomplete" by exigrep. It no longer does this for messages that
4687 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4689 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4690 have allow_domain_literals set, the ID did not get logged in the <= line.
4691 Domain literals are now always recognized in Message-ID: header lines.
4693 4. The first argument for a ${extract expansion item is the key name or field
4694 number. Leading and trailing spaces in this item were not being ignored,
4695 causing some misleading effects.
4697 5. When deliver_drop_privilege was set, single queue runner processes started
4698 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4699 same command in the process it spins off) were not dropping privilege.
4701 6. When the daemon running as "exim" started a queue runner, it always
4702 re-executed Exim in the spun-off process. This is a waste of effort when
4703 deliver_drop_privilege is set. The new process now just calls the
4704 queue-runner function directly.
4710 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4713 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4714 ignored. Though the use of -oP was forcing the writing of a pid file, it
4715 was always written to the default place.
4717 3. If the message "no IP address found for host xxxx" is generated during
4718 incoming verification, it is now followed by identification of the incoming
4719 connection (so you can more easily find what provoked it).
4721 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4723 5. Added some features to "harden" Exim a bit more against certain attacks:
4725 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4726 be put in Local/Makefile. This is like the never_users runtime option,
4727 but it cannot be overridden. The default setting is "root".
4729 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4730 prefix string with which any file named in a -C command line option
4733 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4734 is retained for -C and -D only if the caller of Exim is root. Without
4735 it, the exim user may also use -C and -D and retain privilege.
4737 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4738 command line option is disabled.
4740 6. Macro names set by the -D option must start with an upper case letter, just
4741 like macro names defined in the configuration file.
4743 7. Added "dereference=" facility to LDAP.
4745 8. Two instances of the typo "uknown" in the source files are fixed.
4747 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4748 the Configure-Makefile script screwed up while processing it.
4750 10. Incorporated PCRE 4.4.
4752 11. The SMTP synchronization check was not operating right at the start of an
4753 SMTP session. For example, it could not catch a HELO sent before the client
4754 waited for the greeting. There is now a check for outstanding input at the
4755 point when the greeting is written. Because of the duplex, asynchronous
4756 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4757 way, but not yet received, when the check is performed.
4759 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4760 on TCP/IP sockets, because this apparently causes some broken clients to
4763 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4764 unchanged) from the Cygwin maintainer.
4766 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4767 was supported instead of testing for mbx. Effectively a typo.
4769 15. The spa authenticator server code was not checking that the input it
4770 received was valid base64.
4772 16. The debug output line for the "set" modifier in ACLs was not showing the
4773 name of the variable that was being set.
4775 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4777 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4778 lookup on its own. Something else was needed to trigger the lookup. For
4779 example, a match in host_lookup or the need for a host name in a host list.
4780 Now, if $sender_host_name is referenced and the host name has not yet been
4781 looked up, a lookup is performed. If the lookup fails, the variable remains
4782 empty, and $host_lookup_failed is set to "1".
4784 19. Added "eqi" as a case-independent comparison operator.
4786 20. The saslauthd authentication condition could segfault if neither service
4787 nor realm was specified.
4789 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4790 error message that was logged was misleading, and incoming SMTP
4791 connections were dropped. The message is now more accurate, and temporary
4792 errors are given to SMTP connections.
4794 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4795 (except RSET) with a 421 error, until QUIT is received. However, it was
4796 failing to send a response to QUIT.
4798 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4799 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4800 ACL is now run after the helo_try_verify_hosts code.
4802 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4803 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4804 case-independent, so other case variants are also recognized). Apparently
4805 some systems use these upper case variants.
4807 25. If more than two messages were waiting for the same host, and a transport
4808 filter was specified for the transport, Exim sent two messages over the
4809 same TCP/IP connection, and then failed with "socket operation on non-
4810 socket" when it tried to send the third.
4812 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4814 27. The extern definition of crypt16() in expand.c was not being excluded when
4815 the OS had its own crypt16() function.
4817 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4818 as a preferred synonym for return_size_limit, both as an option and as an
4821 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4823 30. Changed the default configuration ACL to relax the local part checking rule
4824 for addresses that are not in any local domains. For these addresses,
4825 slashes and pipe symbols are allowed within local parts, but the sequence
4826 /../ is explicitly forbidden.
4828 31. SPA server authentication was not clearing the challenge buffer before
4831 32. log_message in a "warn" ACL statement was writing to the reject log as
4832 well as to the main log, which contradicts the documentation and doesn't
4833 seem right (because no rejection is happening). So I have stopped it.
4835 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4836 However, I am unable to do any testing of this.
4838 34. Fixed an infelicity in the appendfile transport. When checking directories
4839 for a mailbox, to see if any needed to be created, it was accidentally
4840 using path names with one or more superfluous leading slashes; tracing
4841 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4843 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4844 contents are added to the log line that is written for every discarded
4845 recipient. (Previously a log_message setting was ignored.)
4847 36. The ${quote: operator now quotes the string if it is empty.
4849 37. The install script runs exim in order to find its version number. If for
4850 some reason other than non-existence or emptiness, which it checks, it
4851 could not run './exim', it was installing it with an empty version number,
4852 i.e. as "exim-". This error state is now caught, and the installation is
4855 38. An argument was missing from the function that creates an error message
4856 when Exim fails to connect to the socket for saslauthd authentication.
4857 This could cause Exim to crash, or give a corrupted message.
4859 39. Added isip, isip4, and isip6 to ${if conditions.
4861 40. The ACL variables $acl_xx are now saved with the message, and can be
4862 accessed later in routers, transports, and filters.
4864 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4865 strings in the file are not string-expanded.
4867 42. If a MAIL command specified a SIZE value that was too large to fit into an
4868 int variable, the check against message_size_limit failed. Such values are
4869 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4870 day this will have to be increased, but I don't think I want to be around
4871 when emails are that large.
4878 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4879 iconv() is not standard in FreeBSD.
4881 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4882 IPv6 enabled. The observed symptom was a segmentation fault on return from
4883 the function os_common_find_running_interfaces() in src/os.c.
4885 3. In the check_special_case() function in daemon.c I had used "errno" as an
4886 argument name, which causes warnings on some systems. This was basically a
4887 typo, since it was named "eno" in the comments!
4889 4. The code that waits for the clock to tick (at a resolution of some fraction
4890 of a second) so as to ensure message-id uniqueness was always waiting for
4891 at least one whole tick, when it could have waited for less. [This is
4892 almost certainly not relevant at current processor speeds, where it is
4893 unlikely to ever wait at all. But we try to future-proof.]
4895 5. The function that sleeps for a time interval that includes fractions of a
4896 second contained a race. It did not block SIGALRM between setting the
4897 timer, and suspending (a couple of lines later). If the interval was short
4898 and the sigsuspend() was delayed until after it had expired, the suspension
4899 never ended. On busy systems this could lead to processes getting stuck for
4902 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4903 process, before it forks any delivery processes. The open lookup caching
4904 mechanism meant that the open file or database connection was passed into
4905 the delivery process. The problem was that delivery processes always tidy
4906 up cached lookup data. This could cause a problem for the next delivery
4907 process started by the queue runner, because the external queue runner
4908 process does not know about the closure. So the next delivery process
4909 still has data in the lookup cache. In the case of a file lookup, there was
4910 no problem because closing a file descriptor in a subprocess doesn't affect
4911 the parent. However, if the lookup was caching a connection to a database,
4912 the connection was closed, and the second delivery process was likely to
4913 see errors such as "PGSQL: query failed: server closed the connection
4914 unexpectedly". The problem has been fixed by closing all cached lookups
4915 in a queue runner before running a delivery process.
4917 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4918 seem to have the "const" qualifier which it has on other OS. I've
4921 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4922 *of the same type* (client or server) with the same public name that an
4923 error should be diagnosed.
4925 9. When Exim looked up a host name for an IP address, but failed to find the
4926 original IP address when looking up the host name (a safety check), it
4927 output the message "<ip address> does not match any IP for NULL", which was
4928 confusing, to say the least. The bug was that the host name should have
4929 appeared instead of "NULL".
4931 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4932 user that is built into the binary, and the -C or -D options is used, root
4933 privilege is dropped before the configuration file is read. In addition,
4934 logging is switched to stderr instead of the normal log files. If the
4935 configuration then re-defines the Exim user, the unprivileged environment
4936 is probably not what is expected, so Exim logs a panic warning message (but
4939 However, if deliver_drop_privilege is set, the unprivileged state may well
4940 be exactly what is intended, so the warning has been cut out in that case,
4941 and Exim is allowed to try to write to its normal log files.
4947 1. smtp_return_error_details was not giving details for temporary sender
4948 or receiver verification errors.
4950 2. Diagnose a configuration error if two authenticators have the same public
4953 3. Exim used not to create the message log file for a message until the first
4954 delivery attempt. This could be confusing when incoming messages were held
4955 for policy or load reasons. The message log file is now created at the time
4956 the message is received, and an initial "Received" line is written to it.
4958 4. The automatically generated man page for command line options had a minor
4959 bug that caused no ill effects; however, a more serious problem was that
4960 the procedure for building the man page automatically didn't always
4961 operate. Consequently, release 4.20 contains an out-of-date version. This
4962 shouldn't happen again.
4964 5. When building Exim with embedded Perl support, the script that builds the
4965 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4966 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4968 6. The freeze_tell option was not being used for messages that were frozen on
4969 arrival, either by an ACL or by local_scan().
4971 7. Added the smtp_incomplete_transaction log selector.
4973 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4974 was accepting AUTH without a new EHLO.
4976 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4977 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4980 10. Logging of TCP/IP connections (when configured) now happens in the main
4981 daemon process instead of the child process, so that the TCP/IP connection
4982 count is more accurate (but it can never be perfect).
4984 11. The use of "drop" in a nested ACL was not being handled correctly in the
4985 outer ACL. Now, if condition failure induced by the nested "drop" causes
4986 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4987 or "require"), the connection is dropped.
4989 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4990 that yield "discard" can now be used with an "accept" or a "discard" verb,
4991 but an error is generated for any others (because I can't see a useful way
4992 to define what should happen).
4994 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4995 are now processed in the same way as lines in the Exim configuration file.
4996 In particular, continuation lines are supported.
4998 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5000 15. Added -ti meaning -t -i.
5002 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5003 domains, and warn by logging if others are found.
5005 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5006 returned from GIFCONF when Exim is trying to find the list of interfaces on
5007 a host. The code in os.c has been modified to copy each ifreq to an aligned
5008 structure in all cases.
5010 Also, in some cases, the returned ifreq's were being copied to a 'struct
5011 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5012 means the last couple of bytes of an IPv6 address could be chopped if the
5013 ifreq contained only a normal sockaddr (14 bytes storage).
5015 18. Named domain lists were not supported in the hosts_treat_as_local option.
5016 An entry such as +xxxx was not recognized, and was treated as a literal
5019 19. Ensure that header lines added by a DATA ACL are included in the reject log
5020 if the ACL subsequently rejects the message.
5022 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5023 MD5 (which is deprecated).
5025 21. When testing a filter file using -bf, Exim was writing a message when it
5026 took the sender from a "From " line in the message, but it was not doing so
5027 when it took $return_path from a Return-Path: header line. It now does.
5029 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5030 with a valid header line field name (a series of printing characters
5031 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5033 23. Changed "disc" in the source to "disk" to conform to the documentation and
5034 the book and for uniformity.
5036 24. Ignore Sendmail's -Ooption=value command line item.
5038 25. When execve() failed while trying to run a command in a pipe transport,
5039 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5040 could be confused with a return value of 69 from the command itself. This
5041 has been changed to 127, the value the shell returns if it is asked to run
5042 a non-existent command. The wording for the related log line suggests a
5043 non-existent command as the problem.
5045 26. If received_header_text expands to an empty string, do not add a Received:
5046 header line to the message. (Well, it adds a token one on the spool, but
5047 marks it "old" so that it doesn't get used or transmitted.)
5049 27. Installed eximstats 1.28 (addition of -nt option).
5051 28. There was no check for failure on the call to getsockname() in the daemon
5052 code. This can fail if there is a shortage of resources on the system, with
5053 ENOMEM, for example. A temporary error is now given on failure.
5055 29. Contrary to the C standard, it seems that in some environments, the
5056 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5057 program. Exim now does this explicitly; it affects the formatting of
5058 timestamps using strftime().
5060 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5061 complaints. I've now initialized all the variables, to avoid this.
5063 32. Header lines added by a system filter were not being "seen" during
5064 transport-time rewrites.
5066 33. The info_callback() function passed to OpenSSL is set up with type void
5067 (*)(SSL *, int, int), as described somewhere. However, when calling the
5068 function (actually a macro) that sets it up, the type void(*)() is
5069 expected. I've put in a cast to prevent warnings from picky compilers.
5071 34. If a DNS black list lookup found a CNAME record, but there were no A
5072 records associated with the domain it pointed at, Exim crashed.
5074 35. If a DNS black list lookup returned more than one A record, Exim ignored
5075 all but the first. It now scans all returned addresses if a particular IP
5076 value is being sought. In this situation, the contents of the
5077 $dnslist_value variable are a list of all the addresses, separated by a
5080 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5081 to accept a host name and all its aliases if the forward lookup for any of
5082 them yielded the IP address of the incoming connection. Now it accepts only
5083 those names whose forward lookup yields the correct IP address. Any other
5084 names are discarded. This closes a loophole whereby a rogue DNS
5085 administrator could create reverse DNS records to break through a
5086 wildcarded host restriction in an ACL.
5088 37. If a user filter or a system filter that ran in a subprocess used any of
5089 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5090 the wrong values were passed to the pipe command ($thisaddress had the
5091 value of $0, $0 had the value of $1, etc). This bug was introduced by
5092 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5094 38. Improved the line breaking for long SMTP error messages from ACLs.
5095 Previously, if there was no break point between 40 and 75 characters, Exim
5096 left the rest of the message alone. Two changes have been made: (a) I've
5097 reduced the minimum length to 35 characters; (b) if it can't find a break
5098 point between 35 and 75 characters, it looks ahead and uses the first one
5099 that it finds. This may give the occasional overlong line, but at least the
5100 remaining text gets split now.
5102 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5103 file descriptors might be low, and that setting 1000 would always raise it.
5104 It turns out that in some environments, the limit is already over 1000 and
5105 that lowering it causes trouble. So now Exim takes care not to decrease it.
5107 40. When delivering a message, the value of $return_path is set to $sender_
5108 address at the start of routing (routers may change the value). By an
5109 oversight, this default was not being set up when an address was tested by
5110 -bt or -bv, which affected the outcome if any router or filter referred to
5113 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5114 log only when "message" or "log_message" are set. However, if one of the
5115 conditions was an address verification, or a call to a nested ACL, the
5116 messages generated by the underlying test were being passed through. This
5117 no longer happens. The underlying message is available in $acl_verify_
5118 message for both "message" and "log_message" expansions, so it can be
5119 passed through if needed.
5121 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5122 new expansion $bh_ to give the encoded byte string without charset
5123 translation. Translation happens only if iconv() is available; HAVE_ICONV
5124 indicates this at build time. HEADERS_CHARSET gives the charset to
5125 translate to; headers_charset can change it in the configuration, and
5126 "headers charset" can change it in an individual filter file.
5128 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5129 that creates RFC 2047 encoded "words" labels them as that charset instead
5130 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5131 expansion operator; (ii) when Exim creates a From: line for a local
5132 message; (iii) when a header line is rewritten to include a "phrase" part.
5134 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5135 buggy, causing it to skip the first lines of messages whose message ID
5136 ended in 'D'. This would not have bitten before Exim release 4.14, because
5137 message IDs were unlikely to end in 'D' before then. The effect was to have
5138 incorrect size information for certain domains.
5140 45. #include "config.h" was missing at the start of the crypt16.c module. This
5141 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5144 46. If there was a timeout during a "random" callout check, Exim treated it as
5145 a failure of the random address, and carried on sending RSET and the real
5146 address. If the delay was just some slowness somewhere, the response to the
5147 original RCPT would be taken as a response to RSET and so on, causing
5148 mayhem of various kinds.
5150 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5151 when I implemented it. It didn't allow for the fact that some option values
5152 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5153 the right test for positive values.
5155 48. Domain names in DNS records are case-independent. Exim always looks them up
5156 in lower case. Some resolvers return domain names in exactly the case they
5157 appear in the zone file, that is, they may contain uppercase letters. Not
5158 all resolvers do this - some return always lower case. Exim was treating a
5159 change of case by a resolver as a change of domain, similar to a widening
5160 of a domain abbreviation. This triggered its re-routing code and so it was
5161 trying to route what was effectively the same domain again. This normally
5162 caused routing to fail (because the router wouldn't handle the domain
5163 twice). Now Exim checks for this case specially, and just changes the
5164 casing of the domain that it ultimately uses when it transmits the message
5167 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5170 50. If a filter generated a file delivery with a non-absolute name (possible if
5171 no home directory exists for the router), the forbid_file option was not
5174 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5175 the existing equality matching.
5177 52. Exim was using ints instead of ino_t variables in some places where it was
5178 dealing with inode numbers.
5180 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5181 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5182 called TMPDIR, and if it finds it is different, it changes its value.
5184 54. The smtp_printf() function is now made available to local_scan() so
5185 additional output lines can be written before returning. There is also an
5186 smtp_fflush() function to enable the detection of a dropped connection.
5187 The variables smtp_input and smtp_batched_input are exported to
5190 55. Changed the default runtime configuration: the message "Unknown user"
5191 has been removed from the ACL, and instead placed on the localuser router,
5192 using the cannot_route_message feature. This means that any verification
5193 failures that generate their own messages won't get overridden. Similarly,
5194 the "Unrouteable address" message that was in the ACL for unverifiable
5195 relay addresses has also been removed.
5197 56. Added hosts_avoid_esmtp to the smtp transport.
5199 57. The exicyclog script was not checking for the esoteric option
5200 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5201 will work only if exicyclog is run under the appropriate euid.
5203 58. Following a discussion on the list, the rules by which Exim recognises line
5204 endings on incoming messages have been changed. The -dropcr and drop_cr
5205 options are now no-ops, retained only for backwards compatibility. The
5206 following line terminators are recognized: LF CRLF CR. However, special
5207 processing applies to CR:
5209 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5210 nor a local message in the state where . is a terminator.
5212 (ii) If a bare CR is encountered in a header line, an extra space is added
5213 after the line terminator so as not to end the header. The reasoning
5214 behind this is that bare CRs in header lines are most likely either
5215 to be mistakes, or people trying to play silly games.
5217 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5218 was being incorrectly given as 18 bytes larger than it should have been.
5219 This is a VOB (very old bug).
5221 60. This may never have affected anything current, but just in case it has:
5222 When the local host is found other than at the start of a list of hosts,
5223 the local host, those with the same MX, and any that follow, are discarded.
5224 When the list in question was part of a longer list of hosts, the following
5225 hosts (not currently being processed) were also being discarded. This no
5226 longer happens. I'm not sure if this situation could ever has previously
5229 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5232 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5233 header. This is something that is recommended in a new Internet Draft, and
5234 is something that is documented as being done by Sendmail. There are two
5235 possible values. For messages generated by the autoreply transport, Exim
5238 Auto-Submitted: auto-replied
5240 whereas for all other generated messages (e.g. bounces) it adds
5242 Auto-Submitted: auto-generated
5244 63. The "personal" condition in filters now includes a test for the
5245 Auto-Submitted: header. If it contains the string "auto-" the message it
5246 not considered personal.
5248 64. Added rcpt_include_affixes as a generic transport option.
5250 65. Added queue_only_override (default true).
5252 66. Added the syslog_duplication option.
5254 67. If what should have been the first header line of a message consisted of
5255 a space followed by a colon, Exim was mis-interpreting it as a header line.
5256 It isn't of course - it is syntactically invalid and should therefore be
5257 treated as the start of the message body. The misbehaviour could have
5258 caused a number of strange effects, including loss of data in subsequent
5259 header lines, and spool format errors.
5261 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5262 client host had authenticated. This control can now be exercised by an ACL
5263 for more flexibility.
5265 69. By default, callouts do not happen when testing with -bh. There is now a
5266 variant, -bhc, which does actually run the callout code, including
5267 consulting and updating the callout cache.
5269 70. Added support for saslauthd authentication, courtesy of Alexander
5272 71. If statvfs() failed on the spool or log directories while checking their
5273 size for availability, Exim confusingly gave the error "space shortage".
5274 Furthermore, in debugging mode it crashed with a floating point exception.
5275 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5276 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5277 serious problem, Exim now writes to the main and panic logs when this
5278 happens, with details of the failure. It then refuses to accept the
5279 incoming message, giving the message "spool directory problem" or "log
5280 directory problem" with a 421 code for SMTP messages.
5282 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5283 0, 1, and 2 exist, because some OS complain for execs without them (see
5284 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5285 descriptors. However, the code omitted to check that the open succeeded,
5286 causing mysterious errors if for some reason the permissions on /dev/null
5287 got screwed. Now Exim writes a message to the main and panic logs, and
5288 bombs out if it can't open /dev/null.
5290 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5291 interact so that it is all more flexible. It is supposed to remain
5292 backwards compatible. Also added extra_local_interfaces.
5294 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5295 to bomb out with an assertion failure - to the client this appears as a
5296 connection drop. This problem occurs in the part of the code that was taken
5297 from the Samba project. Fortunately, the assertion is in a very simple
5298 function, so I have fixed this by reproducing the function inline in the
5299 one place where it is called, and arranging for authentication to fail
5300 instead of killing the process with assert().
5302 75. The SPA client code was not working when the server requested OEM rather
5303 than Unicode encoding.
5305 76. Added code to make require_files with a specific uid setting more usable in
5306 the case where statting the file as root fails - usually a non-root-mounted
5307 NFS file system. When this happens and the failure is EACCES, Exim now
5308 forks a subprocess and does the per-uid checking as the relevant uid.
5310 77. Added process_log_path.
5312 78. If log_file_path was not explicitly set, a setting of check_log_space or
5313 check_log_inodes was ignored.
5315 79. If a space check for the spool or log partitions fails, the incident is now
5316 logged. Of course, in the latter case the data may get lost...
5318 80. Added the %p formatting code to string_format() so that it can be used to
5319 print addresses in debug_print(). Adjusted all the address printing in the
5320 debugging in store.c to use %p rather than %d.
5322 81. There was a concern that a line of code in smtp_in.c could overflow a
5323 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5324 initially expressed, the concern was not well-founded, because trailing
5325 spaces are removed early. However, if the trailing spaces were followed by
5326 a NULL, they did not get removed, so the overflow was possible. Two fixes
5329 (a) I re-wrote the offending code in a cleaner fashion.
5330 (b) If an incoming SMTP command contains a NULL character, it is rejected
5333 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5334 runs initgroups(), so that if the Exim user is in any additional groups,
5335 they will be used during message reception.
5341 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob